Nonextensive statistics, characterized by a nonextensive parameter q, is a promising and practically useful generalization of the Boltzmann statistics to describe power-law behaviors from physical and social observations. We here explore the unevenness of the first-digit distribution of nonextensive statistics analytically and numerically. We find that the first-digit distribution follows Benford's law and fluctuates slightly in a periodical manner with respect to the logarithm of the temperature. The fluctuation decreases when q increases, and the result converges to Benford's law exactly as q approaches 2. The relevant regularities between nonextensive statistics and Benford's law are also presented and discussed.