48

It's common knowledge that mice love cheese. Do they, or is it just that they'll eat anything that's available? What's the origin of the mouse-cheese belief?

  • 15
    Anecdotal evidence: when the building I was unfortunate enough to rent an apartment in had a mice infestation, our success rate wa the opposite of the experiment in the answer: 90% of mice were caught with chocolate; 10% with bread, 0% with cheese. This is when randomly distributing all 3 baits between ~15 traps around apartment for a period of a couple of monthe every day; with >50 mice caught. Randomness was achieved by a Perl script and results were kept in a spreadsheet - you may commence ridicule now :) – user5341 Jun 18 '11 at 17:16
  • 2
    @DVK: Sorry for the delay in the ridicule. Perl and a spreadsheet? Surely R, SPSS, or even Python + Numpy would have been better technologies! :-) – Oddthinking Feb 7 '12 at 13:20
  • 2
    @Oddthinking - to quote a template answer from StackOverflow when a newbie asks "What is the best language for $TRIVIAL_TASK_X": "The best language is one that you already know best or wish to learn as part of this project". In my case, Excel and Perl are what I do best (I'm #11 answerer on StackOverflow Perl tag - used to be #4 before I took a sabbatical :) You may now commence removing tongue from cheek. – user5341 Feb 7 '12 at 15:08
  • 2
    peanut butter always seemed to work better than cheese for me in mouse traps. – JD Isaacks Feb 7 '12 at 18:51
  • 3
    @DVK, tongue removed. It isn't quite as frequent these days, but I have had periods where Excel was my hammer, and everything looked like a spreadsheet. It's remarkable what it can be used for, especially if it is a personal project. – Oddthinking Feb 8 '12 at 1:00
13

I wasn't happy with the lack of peer-reviewed data in the [previously accepted, now deleted] answer, so I did a literature search.

It appears the topic has been examined several times over the years.

They found mice and Norway Rats on New Zealand islands prefer chocolate over cheese.

Of five different bait types tested for monitoring rodent presence, the preferred order was chocolate, cheese, soap, wax and oiled wood.

A number of other studies are relevant, but hidden behind paywalls, such as this one which compared:

cheese, chicken, tuna fish, peanut butter, canary seed, Cat stars, wheat, PCD (MOD) pellets and Non-tox)

However, their results are hidden behind a paywall.

This one, on the other hand published their results in the abstract, but not their methods. They concluded that:

soft wheat, canary seed and rice were the most preferred foods, particularly when fish meal was added.

Although Google says the word "cheese" appears inside the article, it isn't clear whether cheese was tested.


It appears that, while mice will go for cheese, it isn't the irresistable delicacy it is traditionally thought to be, with many other baits preferred.

  • Note that the 2nd study compares a specific strain of mice (West Midlands behaviourally resistant, WMBR) with normal mice. Although WMBR seem to strongly prefer cheese, chicken and tuna to the other types of food, that does not seem to be the case for "normal" mice, which seem not to have very specific preferences in most cases. – nico Feb 7 '12 at 16:57
  • @nico: You have access to the results? I was hoping the preferences of the control mice would be listed - particularly comparing cheese to wheat... – Oddthinking Feb 7 '12 at 20:27
  • The CSIRO study (last one mentioned) tested, amongst other things, wheat flavoured with 0.2% cheese or with 0.2% blue vein. The mice preferred blue-vein-flavoured wheat or plain wheat (or, for that matter, marzipan-flavoured wheat) to cheese-flavoured wheat, but not at a statistically significant level. The experimenters were interested in preferences for taking poisoned bait (but it's relatively moot as the mice were just in a pen selecting from amongst food containers). I'll upload full text later. – Joel Rein Feb 7 '12 at 23:29
  • @Oddthinking: yes they are listed, and they slightly vary depending on the protocol used... but essentially there seems to be no difference between the two in control mice (sometimes cheese is slightly higher, sometimes slightly lower, with no statistical difference). – nico Feb 8 '12 at 7:48
  • 1
    @Sonny: The real history: Oliver C posted an answer based on a quote by a credentialled academic. Commenters complained they couldn't find an article. I found the academic had a history of unscientific pronouncements to the press. I was provoked to write this answer. Oliver, to his great credit, decided to delete his own accepted and highly-upvoted answer. He couldn't (it was accepted) and asked me for mod help. Before I read that, dancek changed his accepted answer (mods can't do that), which allowed Oliver to delete it himself. Integrity all around, and no mod powers used. – Oddthinking Feb 9 '12 at 0:34

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .