Background The Avahan HIV prevention intervention programme promotes condom use amongst men who have sex with men (MSM) in high HIV prevalence states in India. We assessed how self-reported condom use varies with intervention exposure for MSM in Bangalore.

Methods Self-reported condom use and intervention exposure data came from a 2006 cross-sectional survey of MSM. Consistent condom use (CCU) with all, main, and casual male sex partners was assessed. Intervention exposure was measured by: whether MSM had been contacted by Avahan; duration since first contact; number of contacts in the past month; number of condoms received (log-transformed, as plots suggested a non-linear relationship); and number of condom demonstrations seen in the past month. Logistic regression was used to assess the relationship between each exposure variable and CCU with each partner type independently, controlling for demographic, socio-economic and behavioural factors associated with condom use.

Results 77% of MSM contacted by Avahan reported CCU with all partners, compared with 55% of those not contacted (p = 0.019, adjusted analysis). CCU was not associated with duration since first intervention contact or with number of contacts by the intervention in the past month. CCU with all partners and casual (but not main) partners increased with the number of condoms received. CCU increased with the number of condom demonstrations seen in the last month, for all, main and casual partners (odds of CCU with each partner type increased by 1.7–2.4-fold for each additional condom demonstration, p < 0.001 overall).

Conclusion Direct contact with the Avahan programme is associated with increased reported condom use amongst MSM in Bangalore. Although there is potential for social desirability bias, higher levels of reported condom use are associated with repeated contacts with the programme, particularly those involving demonstrations of correct condom use, and with receiving large numbers of condoms from the programme.