The peak-interval procedure is an effective tool for measuring interval-timing behavior in rats. However, its stability has not been previously examined. Therefore, the purpose of this study was to examine behavioral stability across sessions when rats were trained using the peak-interval procedure. To this end, we performed 100 sessions of the peak-interval 30-s schedule with rats. Responses during 90-s probe trials were classified into 3-s bins individually and were averaged across five session-blocks in order to examine behavioral stability between sessions. Data analysis revealed the stability with a relative response rate for one block in five sessions. Furthermore, we applied the provided relative response rate to a normal distribution curve and calculated the kurtosis of the curve at each response rate and Gaussian fitting, i.e., the peak time of the curve. Examination of the stability showed a Cronbach's coefficient of α≧0.86 (<i>R</i><sup>2</sup>=0.98, peak time=0.93, kurtosis=0.86). This indicates that the peak-interval procedure is an excellent reliable tool for measuring the interval-timing behavior in rats.