There are two parts to this paper. The first part examines the impact of arms transfers on the conflictual behavior of third world recipients. I conceptualize conflict as a multiplicative function of total arms imports and the extent to which the recipient is dependent on relatively few arms suppliers. My hypothesis that arms imports encouraged belligerence but that arms-transfer dependence diminished this effect is not widely supported by my time series analyses: only twelve of 86 countries analyzed exhibit this dual pattern. The second part of the paper examines the impact of arms transfers on the aggregate level of military conflict within regional security complexes in the third world. Here I look at both the total amount of arms flows into the region and those arms flows originating with the United States and Soviet Union specifically. Structural hypotheses, which predict the impact of arms transfers based on the characteristics of the regional security complexes, do not receive support from my empirical analysis. Hypotheses that predict regional outcomes based on the source of weapons transfers ? US or Soviet ? fare better. The empirical patterns are consistent with the notion that Soviet arms transfers, representing a flow of military resources and implied political support from a revisionist power, were more destabilizing than arms transfers from the United States, a status quo oriented power