In this article a hypothesis proposed and tested by RUMMEL is taken up again: external and internal conflict are unrelated to each other. In our attempt at replication we use data for 125 nations and for the years 1966-67. Like RUMMEL we use exploratory factor analysis and regression, and, in addition, confirmatory factor analysis. Results from exploratory factor analysis seem to support RUMMEL’s conclusions. Yet confirmatory factor analysis reveals at least moderately strong relationships between internal and external conflict. Regression analysis and partial correlations show, however, that zero order cross-country correlations between internal and external conflict can both be accounted for by population size.