As many as half of all women the world over report having experienced workplace sexual harassment at some stage in their careers. In contrast to existing work focused upon the impact that this has on women, we investigate the impact on firms. First, we model the sexual harassment process at the firm level, showing that the incidence of sexual harassment is jointly determined with female share and firm size. Second, we introduce sexual harassment into the firm’s production function as a drag on female labour productivity, showing that it incentivizes firms to segregate the workforce by gender. Third, we show that impacts of a policy designed to encourage reporting and redressal of sexual harassment will depend upon the elasticity of substitution between male and female labour and TFP and, as a result, policy impacts will tend to vary with baseline female share and firm size. India implemented such a policy in 2013. Using firm-level panel data supplemented with survey data, we estimate reduced form impacts of the policy on firm size and female share. We then calibrate the model, allowing firm-specific production function parameters. We produce estimates of the productivity effects of sexual harassment, and of changes in sexual harassment after the policy as a function of policy parameters.