We use the China shock in Brazil as a quasi-natural experiment to revisit the impact of trade on firm productivity, innovation, and employment. The results corroborate some of the key findings of trade liberalization literature of the 1990s, pointing to a positive although modest effect of trade on productivity. They also point to relatively modest job losses. They raise questions, though, about the effects on innovation, contradicting the positive estimates of the 1990s. This mismatch between productivity and innovation questions the ability of trade policy to deliver sustainable productivity growth on its own.