You are working late in the lab and you need to do a transformation, but sod it, you don't have any antibiotic plates on hand. So you go on the hunt to see if there are any secret stashes anywhere in the lab (you can find secret stashes in every lab if you look hard enough).
And lo and behold, come across the ampicillin plates you poured 4 weeks ago then forgot about. But they are old, so how do you know if they will give adequate selection? Should you use them?
Luckily, someone way back in 1970 determined the lifespan of a whole host of antibiotics in plates, and the good news is that they last a lot longer than you probably thought.
Plates containing methicillin, erythromycin, cephalothin, tetracycline, chloramphenicol, kanamycin, streptomycin, polymyxin B, or nalidixic acid should be a-ok even if stored for one month at 4degC.
Ampicillin was shown to have no loss of activity after 1 week but 10% lower activity after 4 weeks at 4°C, which could result in things like satellite colonies, but should be fine for practical purposes. Other antibiotics with reduced activity after 4 weeks included Penicillin G (23% reduced) and nitrofurantoin (17% reduced).
So even ampicillin plates should be fine for 4 weeks – certainly longer than I would have thought – meaning that your late night experiment can be saved by using your stash of old plates after all.