Tissue microstructure characterization is a valuable tool in diagnosis and staging of many diseases. In this study, we propose a photoacoustic Nakagami statistics method to noninvasively evaluate the number density of random microstructure. The Nakagami parameters are acquired by fitting the photoacoustic signal envelope histogram with Nakagami distribution function. Theoretical calculations and phantom experiments demonstrate that the Nakagami shape parameter is only related to the number density of random microstructure and monotonically increases with the number density. Based on this finding, we propose a photoacoustic tomography modality with the imaging contrast of the Nakagami shape parameter. Experiments show that the proposed method can provide more comprehensive and accurate description of tissue microstructure.