Using multiple network comparisons, we determined whether sexually transmitted disease (STD) rates or black-white STD rate differences are related to how HIV prevention agencies interact through exchanges of information, clients, or funds. In 10 North Carolinian counties, we conducted a cross-sectional analysis, using each county's network as the unit of interest. Using data from face-to-face interviews, measures of the density (overall connectedness) and degree centralization (amount of centralized information hubs) were calculated with UCINET. Statistical significance of the associations between county characteristics and STD rates was assessed with exact Wilcoxon rank sum test. Syphilis rates were lower (P = .01) in counties with a high network density, on the basis of exchanges of information or clients. The effect on black-white rate differences was in the same direction but weaker. Where network degree centralization was high, syphilis rates were higher and the black-white rate differences were larger. Gonorrhea rates demonstrated similar but attenuated patterns of association with network characteristics. STD prevention is more effective when all agencies interact with each other (density) but less effective when interactions are predominantly with a few central agencies (degree centralization). The measures of network exchanges employed in this study may be appropriate for future assessment of service provision of funded prevention programs.