/ Food & Drink, Shopping

Is your food lighter than it claims?

Tin with tape measure

Are the grams on the supermarket products you buy accurate? Our snapshot investigation of 467 supermarket products found that some weighed significantly less…

In 2012 we discovered that 80% of the smoked salmon packs we’d measured were underweight. We wanted to broaden our test for 2015, so we asked food fraud expert Professor Chris Elliot to help us investigate underweight food in supermarkets.

His researchers at Queen’s University Belfast weighed 467 food products bought from supermarkets in Northern Ireland and found that 73 were below the recognised margin of error (this ranges from 4.5g on a product weighing 50g or 15g for a 1kg product).

Products that we found weighed less

Products weighing less than the margin of error:

  • 23 out of 32 tins of Heinz Chunky Veg Big Soup.
  • Six out of eight samples of Tesco Finest smoked salmon.
  • 19 out of 31 Green Giant Niblets Original Sweetcorn contained less than the drained weight stated on packaging.
  • Four out of 32 Del Monte Peach Slices in light syrup. Two were very underweight.
underweight foods

We’d expect a small variation in weight, which the margin of error accounts for, but we were surprised by how many samples were below this.

Enforcement for underweight foods

Trading standards officer Paul Ferris couldn’t comment on our research as it wasn’t carried out by accredited weights and measures professionals, but he did say:

‘Discrepancies may be due to natural desiccation of the product – as stated weight relates to the time of packing – inadequate staff training on packing, or incorrectly using weighing equipment.’

It’s important that manufacturers have effective checks in place to make sure you aren’t getting short-changed. And if products are found to be underweight, it’s important that Trading Standards continues its enforcement to help ensure compliance.

What do you think about underweight food? Are you sometimes suspicious that a product is lighter than it claims? Do you ever go to the lengths of getting the scales out to prove a point?


An excellent exposé.

The Tesco Finest smoked salmon is a premium product and priced accordingly at £4.80 per 100 gm. If the margin of error is 9 gm on a 100 gm pack and 75% of the product in the sample is short of even that benchmark that is a serious contravention, or “diddling” to use the technical term for it. I hope that an accredited professional trading standards officer using the correct and certified apparatus will immediately carry out some evidence-quality tests and take enforcement action if the Queen’s University results are sustained. Should look beyond Tesco as well – it could be widespread.

As for tinned produce, that is more perplexing because I should be most surprised if companies like Heinz, Green Giant and Del Monte did not have weighing equipment in their canning lines [but it has to be calibrated and tested correctly, of course, which requires technical competence]. There is the possibility – but unlikely in my view across three separate manufacturers – that the production lines were adjusted to reduce the weight of the contents [while keeping the price the same?] but nobody corrected the pack weight on the can label. I wonder how full the cans were – with small element product like soup and sweetcorn it should be easy to make sure they are all filled up to the declared weight. With the peaches some extra syrup would bring the weight up if another slice could not be squeezed in, but over 6% of the sample were very underweight – two or more slices short of a full tin perhaps. Since the production and packing processes are probably almost entirely automatic there must be some particular explanation for this and I think we should be told.

And yes, I do often use the weighing scales in supermarkets to check the weights of packed products, but it never occurred to me that the canned fruit, soups and vegetables would be deficient. I suppose it doesn’t occur to Tesco, Sainsbury, Asda, et al, to do routine check-weighing of their bought-in products as well since they are being ripped off by the manufacturers and will have insult added to injury when they have to stand up in court. Selling short weight has historically been a very reprehensible offence.

Apart from the obvious short-changing effect of unreliable product weights for shoppers, it distorts the unit pricing comparisons and value for money calculations that the industry and consumer bodies depend upon.

I wonder what will happen as a result of these companies being found to be breaking the law. Will they just be informed of the problem, or will legal action be taken? A typical response to criticism to investigation of problems of this nature seems to be a prompt and polite response stating that as a reputable company we take the matter seriously and will carry out a detailed investigation. Will anything happen or will it be business as usual once the public has forgotten the news story?

I hope we will find that the short measures are as a result of carelessness and failure to maintaning machinery rather than a deliberate attempt to cheat the public.

Please can Prof Elliott have his second ‘t’.

So even Which? short changes us by giving Chris Elliott fewer letters. ;_D

I find it a cop-out that Trading Standards would not comment because Queens were “not accredited”. Are they incapable of accurately measuring 467 products ? Are tinned products losing weight by dessication? Or are they just reluctant to get involved?

These products are from major manufacturers with the technology and quality control to give the amount of product they wish. What would be useful to know is what proportion of the various products were overweight. That would tell us whether their controls are genuinely defective or whether it all works in their favour. Perhaps Prof. Elliott could tell us.

I would like to see the product weight or volume stated being the “minimum”, not subject to a minus tolerance. Then, without having to dig out the regulations, I will know if I am being short-changed. I should get what I pay for.

I agree that giving a minimum weight would be useful. That would encourage us to go to Trading Standards if we are short-changed. I believe that it is illegal for a barman to give you more alcohol than we pay for but I am not aware of other examples.

We still use the term ‘baker’s dozen’, dating from a time when bakers used to give thirteen items when we ordered twelve. Going back to our history books may be the way forward.

I was told that bakers added such things as chalk so they managed to make 13 loaves from the flour for 12. I don’t think we ever left that era.

wavechange, you are in danger of opening old wounds with you bakers dozen – remember the “Is now the time to ditch imperial and go fully metric?” conversation? The quantity should now perhaps be 11.

Selling in quantities of ten makes it easy to work out the price of a single item. Eggs are often sold in packs of ten.

It would help the consumer if products such as the ones illustrated in Shefalee’s introduction were sold in a multiple of 100g. So full marks for the 400g can of Heinz soup but not for the 285g of sweetcorn or 420g of peach slices. It would make sense to move to 300g of sweetcorn and 400g of peach slices. If these were minimum weights, so much the better.

Excellent idea, but perhaps can-size standardisation has something to do with it and the relative densities of different products. As I said earlier, the deficiency tolerance [and breaches of it] make a mockery of the unit pricing system.

Of course you are right about the nature of the product, John, and it would not make sense to sell English mustard in 400g quantities, like canned soup. For mustard, perhaps 50g and 100g jars would be useful preferred sizes.

With some products such as instant coffee, the common sizes are 100g, 200g and 300g irrespective of manufacturer. A confirmed tea drinker might have a 50g jar for the benefit of visitors, but most buyers buy the sizes I mentioned.

For milk, I suggest 0.5, 1 and 2 litres. It’s time to get rid of strange quantities such as 1.135 litres and 2.27 litres.

I go along with that. A modular system would make life simpler. As happens already with some tinned fruit, cans could have the same circumference but be shorter or taller according to the contents and preferred quantities but all be graduated to sensible common units [multiples of 50g]. It’s interesting that when we had cats we noticed that all the cat food tins were the same size but the weights differed quite considerably across different brands making value comparisons tricky [this was before unit pricing was common on shelf edge labels], so accommodating different densities can be done. With goods in other packaging the restrictions of standardised cylindrical cans do not apply: every manufacturer [and brand] can still have their own distinctive shape in glass or plastic but still follow the modular weight/volume/quantity standard. If you look at washing-up liquids at the moment, the variety of contents volumes across the brands is quite bizarre and the actual amounts seem to make no sense at all – I suspect it’s all done to confuse us over special offers which always seem to come in unique and unrelated quantities. Perhaps it would be argued that the concentration rates for different products packed on the same production lines in identically shaped containers have something to do with it but I see little evidence of that in practice. And I hardly dare mention toothpaste tubes which are almost beyond comparison by virtue of the haphazard sizing [and then they’re boxed in such a way that makes it even more difficult to judge visually]. Unit pricing is the shopper’s safety net but, as we know from other Conversations on the relationship between price and pack-size, errors are commonplace because of the complexity. And putting less product in the same size container for the same price has become an important element of the retailers’ black arts [not forgetting that an identical container in a pound shop will not necessarily contain the same quantity as the same product in a different store, which I am sure fools many customers]. On milk, I think I am correct in saying that the smallest bottles are basically still pints but marked with their metric equivalents!

I expect we will have to carry on wishing for preferred sizes of cans and jars, because any attempt to decrease the amount we were buying would not go unnoticed. We might be better to push for minimum weights, as suggested by Malcolm. When goods are weighed out for the customer, the amount is never underweight and sometimes rather generous.

I was pretending not to understand the strange quantities that milk is often sold in, though I don’t pretend to understand why this is still going on in 2015. Bottled beer generally comes in 0.5 litre measures these days and pubs in the north often supply a generous 0.5 litres in quaint glasses marked ‘Pint’.

Perhaps if we all turn up at the supermarket with an electronic kitchen scale and start weighing the packets as we progress round the store. Even though we cannot measure the weight of the contents of a package, the gross weight will let us select the ones that offer the best value for money and the some others that might interest Trading Standards.

You do need to keep an eye on the till operator if they weigh products to price them at the checkout. They don’t always take their hands off the product so it could cost you more.

I remember when you had to weigh and price up your fruit and veg before you got to the till. One machine had a queue of people waiting at it while the other was free. The reason? The word soon got round that one weighed less than the other !!!

On many occasions I have pointed out that a weighing machine has not been levelled, as apparent from the built-in spirit level. No-one has shown the slightest interest in taking action, even when there was a prominent notice pointing out the need for levelling before use.

Fortunately modern electronic machines are capable of sufficiently accurate weighing when not levelled, unlike their mechanical predecessors.

Sophie Gilbert says:
6 February 2015

This all makes me wonder if there is anything (some) manufacturers (and suppliers etc) won’t stoop to. I know think of them as viruses, forever evolving and finding ways to squeeze money out of us as we find vaccines against them. Unfortunately their natural reservoir is inexhaustible, so after each battle won, the war goes on.

I’m so glad Which is there to inform and unite and fight. What on earth would we do without them?

Thanks for the kind words Sophie. Sometimes I think what we’d do without Sophie Gilbert here on Which? Convo! I don’t think we see enough of you 🙂 Would love if you’d upload an avatar like all the other regulars – it makes it easier for us to spot you. You can do so here: https://conversation.which.co.uk/your-account Thanks!

The reasons some contributors don’t use an avatar may be technical.

From the T&Cs: “We encourage everyone to upload personal avatars, but please make sure they do not contain any potentially offensive imagery. Avatars must be 50 x 50 pixels and no larger than 25 Kbs.”

I doubt that these specification mean much to some of our contributors. Avatars on this site are square, which could be a problem if the image you want to use is rectangular, as many photos and images on websites are.

There are various sites that will create avatars from larger images, though I have not used them and cannot give a recommendation.

Some people respect copyright and avoid re-using images from websites. Maybe a link to a source of copyright-free images would be useful.

Cheers, Patrick!

And testing avatar…

…will need to be adjusted, I see…

Looks good 🙂 Let me know if you need a hand.

Patrick, please could you tell me how to upload an avatar. Clicking on the link you gave to Sophie doesn’t work – it just takes me to the hompage. I sent a message via the Contact Us page a week ago to ask for help, but haven’t heard anything – did you (or somebody at Which) receive my message? I’ve tried clicking on both “Log In” and “Register”, but can’t get anywhere with either.

Hi Esther, sorry we didn’t spot your email. It may have got stuck in our spam filter. Before you can access the page I sent to Sophie, you need to be logged in. Have you had trouble logging in? Your particular username happens to be the same as your email address. Once you’re logged in you can then get to https://conversation.which.co.uk/your-account

If you’ve forgotten your password and can’t login, you can use the forgotten password link to reset it. Let me know if that doesn’t work. Thanks

If we bought a box of eggs and one was missing, we might be annoyed enough to take it back to the shop and demand a refund or replacement. We certainly would with more expensive items than eggs.

If a packet is labelled that if contains approximately 20 items, many of us will count them.

I doubt if many of us routinely check the weight of pre-packed food, even if we have kitchen scales that measure to 1 or 2g in our kitchen.

The sellers don’t check them either, obviously, yet they are not getting what they are paying for and will have to defend themselves if it comes to court. Perhaps they think they can just pass the blame over to the manufacturers, and in this crazy world maybe they can. “Call The Man from Del Monte!” [He usually says “YES” but probably not on this occasion.]

On a 100 to 200g pack the margin for error is, i believe, 9%. This means that every pack sold could be 91g or 182g of course – with the wholesaler and retailer making 10% extra on the product. This margin seems quite inequitable to the consumer (and the unwitting retailer perhaps). It would make more sense, given some tolerance must be allowable, to either use a minimum weight (as I suggested earlier) or to make this an average weight, so packs could be from -9% to +9%. Trading Standards could then check a batch of product for compliance.

Incidentally, has any action been taken from the findings of the original investigation? Little point in moaning about these sharp practices if nothing is pursued.

One food that is open to abuse is steak in restaurants.

I once ordered an 8oz steak medium to well done. When it arrived it was tiny. I told the waitress that it was not an 8oz steak but she and the manager insisted the weight had been lost in cooking. I just about got 4 mouthfuls out of it.

Needless to say we never went back.

I wonder if Which? has ever considered involving schools with simple projects that involve little resources. Checking for underweight products would be an ideal example, and schools will already have balances (what most people know as scales) capable of measuring weight accurately.

Younger children would find this fun. Older children could learn more about the scientific issues involved in weighing, some simple and others more challenging.

We need to get children involved with consumer issues from an early age, and encouraging them to think about a career in science would not come amiss.

School projects would be great publicity for Which? and it might encourage parents to get more involved in consumer affairs.

What a great idea wavechange – I’ve passed forward your comments to our Research Teams for their consideration.

Brian Johnson says:
9 February 2015

I am a retired Trading Standards Officer / Weights and Measures Inspector and was trained in statistics in order to enforce the ‘Average Quantity’ legislation which basically says these days that packers and importers must comply with three rules namely:

– the actual contents of the packages should not be less, on average, than the nominal quantity;
– the proportion of packages which are short of the stated quantity by a defined
amount (the “tolerable negative error” or TNE) should be less than a specified level; and
– no package should be short by more than twice the TNE.

I wonder how many compliance checks are carried out by Weights and Measures Authorities (who have a duty to enforce the Regulations (The Weights and Measures (Packaged Goods) Regulations 2006 (SI 2006/659) ) in these times of austerity and local government cut backs – the findings of Queens University suggest that it may not be enough!

Thanks for that information Brian. So far as you were aware from your experience, was there any routine testing of manufacturers’ own weighing equipment in the factory to ensure that correct weight was going in the tins or packets in the first place? Obviously this topic is only a tiny snapshot of a situation and might be completely unrepresentative, but if there are not enough TSO’s not only will there be no enforcement at the customer end of the trade but it will be impossible to place any reliance on the production process getting it right in the firts place. If staff are not being adequately trained to calibrate the machinery or to use it correctly, as has been suggested, then there are bound to be more weight/volume/quantity problems at the point of purchase and, as this example demonstrates, the fluctuations could be quite wide of the mark. Strategically it’s better to put the inspection and testing at the front end to save excessive compliance checking and enforcement at the far end of the supply chain [although spot checks and sampling should still be done proactively in my opinion without having to wait for a customer to complain – which, as soon as they realise not much will be done, they won’t.

Just bought multi bag of crisps from Lidl, SNAKTASTIC and found the last bag considerably underweight it should have been 25g but weighed only 10g.

Hi Bob – I suggest you contact Citizens Advice and see what they advise: https://www.citizensadvice.org.uk

Which? raised this issue earlier in the year: https://conversation.which.co.uk/food-drink/underweight-supermarket-food-cans-tins-weight-grams/

Roger Richmond says:
26 October 2016

Interesting this. I’ve just complained to my local Sainsbury about a pack of pork chipolata sausages labelled at 340 gm, but only weighing 315 gm. This is the third time I have complained about their fresh meat weights in the last 18 months. Is it Sainsbury or their suppliers?

That’s an interesting one, Roger. You have paid Sainsbury so they are responsible for selling an underweight product, so they should deal with your complaint. They in turn can – and should – take it up with their supplier. Weight is very easy to measure now that we have accurate and dependable equipment, so I think it’s time to require suppliers to provide a minimum of the exact weight. We have moved on from the days of inaccurate spring scales that need to be levelled to measure accurately.

You used to buy pre-packed steak in Sainsbury at their actual weight, now they are a set weight. Mostly they are overweight, but we have had one under.

According to an EU directive I have come across (76/21 1 /EEC) prepackaged goods between 300 and 500 (g or ml) must be within 3% of the equantity. So if your pack was marked e340g I’d suggest the minimum you should be given is 340 – 3% = 330g. If so you could complain to Trading Standards if it is a regular occurrence (that is if TS will deal with you) but I’d email Sainsbury’s head office.

I have just bought a frozen goose crown from Lidl, stated weight 1000g but actual weight 914g. I shall be taking it back but from reading the earlier posts I’m not sure of my legal rights to return the item. I can do so under their ‘satisfaction’ policy but otherwise presumably it would fall under the ‘not as described’ category – if this applies to grocery items? All rather unclear.

Hi Mary – From official guidance for business: https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/487018/Guidance_-_The_Weights_and_Measures__Packaged_Goods__Regulations_2006_v.4_December_2015.pdf

The Packers’ Rules
14. The Directive sets out three rules with which packers must comply:
– the actual contents of the packages should not be less, on average, than the nominal quantity;
– the proportion of packages which are short of the stated quantity by a defined amount (the “tolerable negative error” or TNE) should be less than a specified level; and
– no package should be short by more than twice the TNE.

In your case the TNE on 1000g is 1.5% or 15g and no package should be short by more than 30g. At 914g, yours is a Lidl short of the minimum of 970g. I will be interested to see what they have to say about this.

David Farmer says:
6 November 2017

Within the last 3 months I’ve had a 1kg Boswell Farm mince weighing 900g (fm Tesco) and a Tesco frozen 500g Pork Should weighing 360g (421g with all packaging). Thats some difference.

Recently felt that Tesco eggs were getting a lot smaller. At the moment have medium Tesco eggs and having weighed them found that 2 were under the minimum weight and actually fell into the small category. Have eaten the rest so don’t know what they weighed. Will weigh the next lot and if incorrect will get in touch with Trading Standards

Recently I bought a 5kg packet of bean seed which when weighed using the instore weighing scale weighed 4.85kg. Is this normal?

This comment was removed at the request of the user

The regulations are summarised here: https://www.gov.uk/weights-measures-and-packaging-the-law/packaged-goods

For a 5kg pack, the “no package can be underweight by more than twice the tolerable negative error (TNE)” The TNE applicable to 5kg of a product is 150g, so that the shortage is permissible provided that “the contents of the packages must not be less, on average, than the weight on the label”.

If weighed in the store, this will include the packaging, so Jade has a case for taking the matter up with the retailer or, as Duncan suggests, Trading Standards.

It would be interesting to know whether the label on the shelf gave a different weight to that shown on the packaging. I have come across such discrepancies when the manufacturer has reduced the contents weight but not amended the data on the packet whereas the shelf label showed the weight correctly.

One of the most confusing things for consumers today is the frequent changes in weight, quantity and size of well-known products without it being apparent in the packaging. This has become an accepted, but in my view not acceptable, merchandising technique.

jw says:
9 April 2019

Recently I have been buying 150g blueberries from Sainsbury’s and using 50g a day. Every third day I am short. Today I weighed a full packet, including packaging the total was 148g including packaging (7g). That’s about an extra £2 for Sainsbury’s for every 15 sold. Good business for them I guess.

It is always worth weighing things sold by weight in Sainsbury’s. There is usually a machine in the fruit and veg section, just take a couple you prefer, weigh them, and there is every chance you will get your money’s worth.

The three packers’ rules
A batch of packages must, at the time of packing, comply with the following rules:
 the actual contents of the packages must not be less, on average, than the nominal quantity (that declared on the container)
 the proportion of packages that are short of the declared quantity by a defined amount (the ‘tolerable negative error’ or TNE) should be less than a specified level (generally no more than 2.5%)
 no package should be short by more than twice the TNE

Essentially, the average weight must be no less than the nominal. “Average” means some packages will be under nominal. The TNE for a package of 100-200g is 4.5%. Your blueberries are, from what you weigh (depends on your scale accuracy) around 6% under weight, within the limits but pushing them. Have you raised this with Sainsbury’s? They may not realise they might be being short changed by their supplier (although they should do checks)..

I wonder whether it would not be fairer to make all prepackaged food no less than the nominal weight? This would, in practice, mean every package would be overweight to avoid potential prosecution, and that would put the price up by maybe 5-10%. Partly a practical issue about the accuracy in production and shrinkage in store (loss of moisture in fruit for example). Would we want that?

Yes, if the plums have turned into prunes they could be underweight.

I would be happier for weights shown to be the minimum weight, although a publican must not be too generous in their measures.

Jim says:
21 April 2019

I regularly split a 400g tin of custard into two and use scales to get it right. The amounts for each portion consistently weigh 190g. Is 20g lighter every time acceptable?

Jim – This link gives details of what is legally acceptable: https://www.gov.uk/weights-measures-and-packaging-the-law/packaged-goods

The ‘tolerable negative error’ for 400g of product is 3%, which corresponds to 12g, and no package can be deficient in more than twice this, or 24g, so your half measure would correspond to a minimum of 200-12 = 188g, so 190g would be acceptable. If you could prove that the goods were consistently underweight that is certainly not acceptable.

It’s worth checking that your scales are accurate. I know that both my scales are accurate to within 1g, having checked them with accurate weights, but have seen examples where scales under- or over-read.

Denise says:
23 June 2020

I recently purchased a 24 bisk pack of weetabix and there were only 23 in it. I contacted them and was sent a £1 voucher. This has happened before so it makes me wonder how much they could save over a long period

Same with milk. Sainsbury’s milk cartons have a line to say where 4 pints should be, I have looked at recent cartons and every single one is the same place below this line. High tech equipment can fill very accurately and clearly they are working to the allowed margin of error.

Also, toothpaste and moisturiser have a lot of air bubbles in now. Again machines exist to remove/minimise these. Almost as if they do it on purpose!