Faced with contrasting results between women on corporate boards (WOCB) and corporate social performance (CSP), this issue warrants further investigation. Given that the relationship between the two is questionable, semiparametric generalized additive models (GAM) are used herein, making it possible to capture the non-linearity of the WOCB-CSP relationship more effectively. This contribution is significant, as many authors have demonstrated the nonlinearity of factors influencing performance, whether of a financial or social nature. Consistent with token and critical mass theories and based on a sample of firms from Fortune 1000 over the period between 2004-2018, the effect of WOCB on CSP is found to be nonlinear. GAMs were used herein to demonstrate that a minimum of WOCB is necessary to generate CSP. However, a critical mass of WOCB is not synonymous with improved CSP. In so doing, this study sheds light on the relationship between WOCB and CSP, which remains a controversial issue.