Two experiments examined about interval timing behavior in rats. In the first experiment, we examined the effects of the self-start lever on peak procedure. The aminals were divided into the self-start group and the control group. For the self-start group, the animals were required to press the self-start lever to start the trial. For the control group, the trial started automatically independent of responses. Then, the animals were trained using the peak procedure that included fixed interval (FI) trials and probe trials. The FI trials consisted of a food pellet reinforcement that presented by the first lever press response more than 30-s after tone stimulus onset, and the stimulus was terminated. The probe trials delivered no reinforcement with responses and the tone stimulus lasted for 90-s. The temporal distriutions of lever press responses in rats on probe trials were similar to the Gaussian function curve. It had a peak point, which showed the peak of regression curve fitting with the Gaussian function. It corresponded to near the 30-s with reinforcement time. In the self-start group, the number of responses at the beginning of the trials was increased than that for the control group. The temporal distribution of responses for self-start group was slightly distorted. In the second experiment, we examined the combination of the self-start lever and the gap procedure. On the gap procedure, trials were started with the onset of tone stimulus and followed by the offset of tone stimulus during gap interval, and tone stimulus presented again after gap interval. Consequently, both groups approximately kept timing during gap interval. It suggests that the tone stimulus hardly operates as a cue of temporal information in this experiment.