Background and Objective The prevalence of HIV among MSM continues to be disproportionately high in India. The objective of this study was to identify the factors associated with change in the prevalence of HIV testing overtime and to assess their relative contribution in increasing (or decreasing) the prevalence of HIV testing among men who have sex with men.

Methods Data are drawn from two rounds of cross-sectional bio-behavioural surveys of 7403 (round I: 3895, round II: 3508) self-identified MSM from three states of India, recruited through probability based sampling in 2005–06 and 2009–10 respectively. Logistic regression models were used to assess the relationship between socio-demographic characteristics, sexual behaviours, programme exposure and HIV testing. Significant factors were further parsed using decomposition analysis to examine the contribution of different components of that factor towards the change in HIV testing over time.

Results There was a significant increase in the proportion of MSM reporting HIV testing from round I to round II (13% vs. 77%: p < 0.001). Age, literacy, self-identified sexual identity, type of partner and exposure to intervention were significantly associated with HIV testing. Literates; those in the age groups of 25 to 34; kothis (predominantly receptive); those who had both commercial and non-commercial partners and those who were exposed to the intervention contributed the most to the increase in HIV testing over time.

Conclusions These study results underscore the potential of identifying and targeting segments within the at-risk MSM population which are currently not accessing HIV testing services. Voluntary HIV testing affects sexual behaviour leading to a positive motivation for safer sexual practises.