The theory of strategic trade policy yields ambiguous recommendations for assistance to exporting firms in oligopolistic industries. However, some writers have suggested that investment subsidies are a more robust recommendation than export subsidies. We show that, though ambiguous in principle, the case for investment subsidies is reasonably robust in practice. Except when functional forms exhibit arbitrary non-linearities, it holds under both Cournot and Bertrand competition, with either cost-reducing or market-expanding investment, and with or without spillovers. Only if firms have strong asymmetries in their investment behaviour and engage in Bertrand competition is an investment tax clearly justified.