QSAR analysis has been done to correlate antimicrobial activity of substituted benzimidazole derivatives with their physicochemical parameters. Developed QSAR models have been cross validated using leave one out (LOO) method. Statistical parameters like probable error of the coefficient of correlation (PE), least square error (LSE), Friedman's lack of fit measure (LOF), standard error of prediction (SEP) and quality value (Q) were also used to cross validate the models. QSAR studies established the importance of WAP, Mlog P and UI in describing the antimicrobial activities of substituted benzimidazole derivatives.