Two non-reactive field experiments were conducted to study factors influencing stealing. The subjects were people in the streets of London who picked up a stamped, addressed, unsealed, apparently lost letter, containing a handwritten note and in most cases also a sum of money. Experiment 1 employed a 2 x 2 x 2 between-subjects factorial design, varying the amount of money (20p or 1 pound), the form of the money (cash or postal order) and the apparent victim (an old lady or a higher class male). Ten letters were dropped in each of the eight conditions, and a further 10 in each of two control (no money) conditions. Stealing was greatest with cash and the higher class male victim, but did not vary significantly with amount of money. In Expt 2, larger amounts of money were used, and the apparent victim in each case was the old lady. Twenty letters were dropped in each of three conditions, control (no money), 1 pound and 5 pounds in cash, and stealing did vary significantly with amount of money. In both experiments, younger people were more likely to steal, as were those who put the letter in pockets or handbags after picking it up.