While many scholars have posited a simultaneous relationship between trade and conflict, very few empirical studies have specified the relationship as such. Those that did employed samples that were relatively limited in spatial-temporal coverage. None have employed conflict indicators based on Militarized Interstate Dispute (MID) data due to the difficulty of including discrete dependent variables in a simultaneous equation framework. We overcome both limitations in this study and offer results with important theoretical implications. In line with the most recent published studies, we design our model to “condition” the estimates on recent histories of dyadic trade and conflict. We apply Maddala's estimator, which is designed for a two-equation system in which one endogenous variable is continuous and the other is dichotomous. While the signs of all control variable coefficients match those reported in mainstream trade-conflict literature, our main result does not. We find what we call the “primacy of politics”; i.e., that conflict indeed inhibits trade while the effect of dyadic interdependence on the likelihood of conflict is statistically insignificant. Extensive sensitivity analyses show the results to be consistent across estimators as well as alternative model specifications and trade data sets. We conclude that liberal claims about interdependence and conflict may be problematic.