Abstract Background The HIV epidemic in Peru is still regarded as concentrated - sentinel surveillance data shows greatest rates of infection in men who have sex with men, while much lower rates are found in female sex workers and still lower in the general population. Without an appropriate set of preventive interventions, continuing infections could present a challenge to the sustainability of the present programme of universal access to treatment. Determining how specific prevention and care strategies would impact on the health of Peruvians should be key in reshaping the national response. Methods HIV/AIDS prevalence levels for risk groups with sufficient sentinel survey data were estimated. Unit costs were calculated for a series of interventions against HIV/AIDS which were subsequently inputted into a model to assess their ability to reduce infection transmission rates. Interventions included: mass media, voluntary counselling and testing; peer counselling for female sex workers; peer counselling for men who have sex with men; peer education of youth in-school; condom provision; STI treatment; prevention of mother to child transmission; and highly active antiretroviral therapy. Impact was assessed by the ability to reduce rates of transmission and quantified in terms of cost per DALY averted. Results Results of the analysis show that in Peru, the highest levels of HIV prevalence are found in men who have sex with men. Cost effectiveness varied greatly between interventions ranging from peer education of female commercial sex workers at $US 55 up to $US 5,928 (per DALY averted) for prevention of mother to child transmission. Conclusion The results of this work add evidence-based clarity as to which interventions warrant greatest consideration when planning an intervention response to HIV in Peru. Cost effectiveness analysis provides a necessary element of transparency when facing choices about priority setting, particularly when the country plans to amplify its response through new interventions partly funded by the GFATM.