Large aftershocks can cause additional losses and trigger the risk of building damage. To study the decay law of aftershock sequences, this paper attempts to use the exponential decay model to fit and analyze five aftershock sequences in various regions. The performances of the exponential and traditional aftershock decay models were analyzed using the corrected Akaike information criterion, Bayesian information criterion, and adjusted R². The results show that the ability of the exponential model to describe the decay law of aftershock sequences is close to that of the modified Omori aftershock model and the stretched exponential model, especially the M_S6.0 aftershock sequence in Changning, Sichuan Province, and the M_S5.7 aftershock sequence in the Yiliang county of the Yunnan Province. In these two sequences, the exponential model performed better than the other two models, and the exponential model parameters showed clear physical meanings. The sum of parameters A and r can accurately represent the number of initial aftershocks after a strong earthquake. The deviations in the initial aftershock number of the five aftershock sequences are all less than 1.70%. The parameter k can be used as a characteristic value that reflects the decay rates of the aftershock sequences. The larger the value of k, the slower the decay of the aftershock sequence; and its values are inversely proportional to the magnitude of the main shock.