Two experiments examined interval timing behavior in mice and rats. In the first experiment, subjectswere mice, and in the second one they were rats. The animals were trained using the peak procedure that was avariant of a discrete trial fixed interval (DT-FI) schedule. On food trails, the first lever press response more than30-s after tone stimulus onset terminated the stimulus and presented a food pellet. On empty trials, the tonestimulus lasted for 90-s with no food. The temporal distributions of responses in mice and rats were calculatedby averaging responses over 180 empty trials. Training progressed in both animals in a way similar to Gaussianfunction curve. Peak time, which shows the peak of regression curve fitting the Gaussian function, correspondedto 30-s with wide variability in both experiments. This proved to be a good index to the interval timing behavior.In the second experiment, we considered various regression fitting variables, especially concerning the numberof trials. Temporal distribution seems to be affected by various factors including individual differences intraining sessions.