The article presents research on optimizing grinding parameters for abrasive belt grinding of surgical tools in the context of replacing manual processes with automated machining. The main objective was to determine the minimum acceptable surface quality after grinding that would enable achieving the required quality after vibratory finishing. The research was conducted on samples replicating the shape of plaster cutting scissors, using belts with grits P50, P60, and P80 in various grinding strategies. The influence of the number of passes (1, 2, 3) and feed rate (F100, F200, F300) on surface roughness (Ra, Rz) before and after vibratory finishing was analyzed. The conducted ANOVA test for trials 1–9 showed that at a significance level of α = 0.05, for a given longitudinal feed rate, surface roughness does not depend on either belt grit size or the number of passes. This result may stem from the specific characteristics of surgical tool materials and the use of machining parameters atypical for coarse belts. The findings indicate that a grinding strategy with multiple passes at higher feed rates (F200, F300) is less favorable than a strategy with a single pass but lower longitudinal feed rate.