Empirical studies have demonstrated that both firms' exports and imports increase their productivity, although it may have different upstream effects on domestic firms. This study revisits the propagation of trade effects through interfirm transactions by improving the methods of previous empirical analyses in three ways. First, it uses stricter criteria for sampling firms in order to estimate the effects without bias from other international transactions. Second, it deals with the indirect impact of trade shocks on various indices of upstream suppliers, such as the possibility of closure, the number of workers, and productivity. Third, it employs a one-to-one propensity score matching combined with a difference-in-differences approach, a method that controls both buyers' and sellers' characteristics. Results show that there is no systematic trade effect on upstream seller firms, and that most of the trade impacts on business performance variables of seller firms are statistically insignificant. One possible reason is that firms that increase their exports or imports do not sufficiently change their purchase of material and intermediate goods from domestic non-associated firms, a supposition that is supported by the empirical analysis. The result suggests that the economic impact of firms' international trade on upstream suppliers is more nuanced than just a substitute or complement between international and domestic trades.