The purpose of this study was to determine if preservice middle school mathematics teachers' (PSTs) satisfaction levels towards peer instruction (PI) method showed a meaningful difference according to their learning approaches and individual innovativeness characteristics.The data collected from third-year PSTs, who were enrolled at the mathematics education program of a state university in Turkey. The study was conducted in a statistics and probability course with 46 PSTs, who answered all of the questions in the data collection tools. The findings showed that while 33 of the 46 PSTs had a deep learning approach, 13 had a surface learning approach. According to the descriptive analysis, the PSTs' average individual innovation scores was 64.43 and the standard deviation was 7.85. The PSTs' distributions according to the innovation categories were determined as follows: Two of them were innovators, eight of them were early adopters, 32 of them were early majorities, three of them were late majorities, and one of them was laggards. Furthermore, the analysis showed that the PSTs' satisfaction levels towards the PI method did not significantly differ according to their learning approaches and characteristics of innovation.