Social marketing interventions have been shown to both promote and change many health-related behaviours and issues. As the HIV epidemic continues to disproportionately affect MSM and transgender women around the world, social marketing interventions have the potential to increase HIV/STI testing uptake among these populations. To assess the impact of social marketing interventions on HIV/STI testing uptake among men who have sex with men and transgender women compared to pre-intervention or control group testing uptake in the same population. We searched the following electronic databasesfor results from 01 January 1980 to the search date, 14 July 2010: Cochrane Central Register of Controlled Trials (CENTRAL), EMBASE, LILACS (Latin America and Brazil), PsycINFO, PubMed, Web of Science/Web of Social Science, Chinese National Knowledge Infrastructure (CNKI), and CQ VIP (China). We also searched for conference abstracts in the Aegis archive of HIV/AIDS conference abstracts and the CROI and International AIDS Society websites. In addition to searching electronic databases, we searched the following sources of grey literature: Australasian Digital Theses Program, Canadian Evaluation Society, Eastview: China Conference Proceedings, ProQuest Dissertations and Theses, and World Health Organization Library Information System (WHOLIS). We contacted individual researchers, experts working in the field, and authors of major trials for suggestions of any relevant manuscripts that were in preparation or in press. References of published articles from the databases above were searched for additional, pertinent materials. All languages were included in this search. Randomized controlled trials and controlled clinical trials that compared social marketing interventions with a control were included. Interrupted time series and pretest-posttest design studies (controlled or uncontrolled) that compared social marketing interventions with no intervention or a control were also included. Posttest-only studies and studies that combined pre-post data were excluded. Interventions that targeted at general public but did not include MSM or transgender women as a segment or did not have outcome data for an MSM or transgender segment were excluded. Two authors independently extracted data from each included study and assessed study quality. Meta-analyses were conducted to compare pre- and post-intervention and intervention and control group outcomes of HIV and STI testing uptake. Quality of evidence was assessed using the GRADE approach. Three serial, cross-sectional pretest-posttest study designs (one with a control group and two without) were included in the final analysis. Statistical pooling was conducted for two studies and compared to pre-intervention level testing uptake, which showed that multi-media social marketing campaigns had a significant impact on HIV testing uptake (OR=1.58, 95%CI = 1.40 - 1.77). However, the campaigns were not found to be effective in increasing STI testing uptake (OR=0.94, 95%CI = 0.68 - 1.28). Overall, risk of bias was high and quality of evidence was low. None of the studies were conducted in developing countries or included male-to-female transgender women. This review provided limited evidence that multi-media social marketing campaigns can promote HIV testing among MSM in developed countries. Future evaluations of social marketing interventions for MSM should employ more rigorous study designs. Long-term impact evaluations (changes in HIV or STI incidence over time) are also needed. Implementation research, including detailed process evaluation, is needed to identify elements of social marketing interventions that are most effective in reaching the target population and changing behaviours.