Drinking rhythm in 6 rats optically enucleated on the day of birth was determined every one to two weeks after weaning until the age of 28 weeks. Each rat was transferred repeatedly from a cage without a running wheel to one with a running wheel, and vice versa. During the periods when the rats were housed in the former cage, drinking rhythm delayed, while it advanced when they were housed in the latter. The rats, which had a period of free-running rhythm of motor activity longer than 24 hr when measured by an Automex device in a cage without a running wheel, showed a period shorter than 24 hr when measured by a running wheel. These results suggest that the difference in tools for determination of activity may influence the results of animal experiments for free-running rhythm, and warn investigators of risks being caused by the selection of tools.