A method for measuring activity of hamsters using a stabilimeter at a 1 second sampling rate with data computer recorded as 5 minutes integrated values was developed. In a single cage without a running wheel a consistent pattern for activity was observed, consisting of (a) low levels of daytime activity until one or two hours before lights off when activity increased significantly; and (b) a peak of nocturnal activity in the first hour of the dark cycle. The inclusion of a running wheel increased and altered significantly the pattern of nocturnal activity. In further experiments animals were housed in two linked cages, one acting as light-proof burrow and the other exposed to light. Measurements were recorded from each cage independently and from two position detectors in the interconnecting tunnel. The results showed: (a) total activity, i.e., the summation of activity in both cages, was not different from activity in a single cage system; (b) low daytime activity was composed of prolonged periods of rest in the burrow plus short periods of activity in the exposed cage; the increased activity one hour before lights off was localised to the light-proof burrow; and (c) after lights off, the animals began to spend increasing periods of time in the exposed cage reaching a maximum after one hour. Replacing artificial with natural light did not change the principal features of behaviour.