Trade theory postulates that trade liberalization produces gains from trade though rising competition and allocative efficiency, but the available empirical evidence is mixed. This paper tests the theory using Indonesia’s manufacturing firms’ data from 1990 to 2015. The study examines how markup, productivity, and misallocation respond to trade liberalization using changes in effective rate of protection (ERP), output, and input tariffs as measures. To isolate exogenous variation of output tariff, the study uses instrumental variables that capture tariff reforms driven by multilateral liberalization commitment, which is unrelated to domestic lobby-group pressure that impacts unilateral liberalization. Three main findings emerge. First, the effect of trade on markup depends on the type of liberalization. While output tariff liberalization increases competitiveness (lower markup), input tariff liberalization raises markup, suggesting incomplete pass-through effect. These effects mainly work through a within-firm channel. Second, the study finds allocative efficiency gain of trade, where a reduction in ERP enhances manufacturing productivity through reallocation of resources towards the most efficient users. Third, greater exposure to trade liberalization reduces resource misallocation. These results contribute to broadening our understanding of gains from trade under distorted settings and suggest that trade liberalization can reduce misallocation that often adversely impacts manufacturing productivity in developing countries.