We propose and demonstrate a new method for evaluating the afterpulsing effect in single-photon avalanche photodiodes (SPADs). By analyzing the statistical property of dark count rate, we can quantitatively characterize afterpulsing probability (APP) of a SPAD. In experiment, the temperature-dependent low dark count rate (DCR) distribution becomes non-Poissonian at lower temperature and has higher excess bias as the afterpulsing effect becomes significant. Our work provides a flexible way to examine APP in either single-device or circuit level.