Background An increasing body of evidence indicates venues where people recruit sex partners have a role in spread of STIs. Swingers recruit their sex partners by physical venues (clubs) and virtual venues (websites), forming so-called sexual affiliation networks. Objective of the present paper is to examine how these sexual affiliation networks of swingers can be relevant for STI prevention.

Methods Participants of our swinger’s cohort were followed using questionnaires. We used both conventional epidemiological and social network methods (descriptive and Exponential Random Graph Models (ERGM)) for the analysis. Because couples swing together, they were found equal in their choices of venues and taken as a swing unit (SU).

Results The 57 SU frequented 13 (33.3%) clubs and 26 (66.7%) websites; 59.6% (n = 34) of the SU frequented both websites and clubs, 36.8% (n = 21) frequented only websites and 2(3.5%) solely clubs.

The network formed only one component (fragmentation = 0): all SU and venues in the network were interconnected. The sexual affiliation network has a clear core-periphery structure, with a core of highly connected SU (n = 16) and venues (n = 7).

SU characteristics were generally not statistically significantly associated with the number of websites or clubs frequented, except for the following three: swinging years, group sex, and drug use.

Drug users had a significant stronger tendency to visit websites than non-drug users, and also had a strong tendency to frequent websites more than clubs.

Conclusions The analysis of sexual affiliation networks has an important added value to conventional STI epidemiology, because the last does not assess the direct relations between actors and mechanisms of clustering. Furthermore our results show that all swingers in the network were interconnected through their affiliations and therefore STI can eventually reach everyone. Interventions that focus on the prevention of drug use should be directed to visitors of websites instead of clubs.