This study investigates the effect of the economy structure on the U.S. - China bilateral trade deficit as alternative to the influence of the exchange rate fluctuation. The revealed comparative advantage indices are proposed as the measure of the relative structural differences between two countries due to factor endowments and technology. A Bayesian Stochastic Search Variable Selection method is applied to the U.S. - China annual trade data for 57 commodity groups at the SITC 2-digit industry aggregation level to obtain empirical variable inclusion probabilities. Based on the data, we found no conclusive evidence against the hypothesis of the short-run effect of either of the explanatory factors, while the long-run influence is revealed to be insignificant in most of the cases.