Objective: To estimate the impact of antiretroviral therapy in reducing the number of new HIV infections under different intervention strategies targeting men who have sex with men in Colombia. Methods: We constructed a mathematical model for the transmission of HIV. The structure reflects the natural history of disease and initiation of antiretroviral treatment according to different levels of CD4 cell count. A review of local sources informed the values for behavioural parameters and interventions in the model. Calibration was performed using a random sampling strategy within the parameter sample space followed by a rejection algorithm based on minimum acceptance limits according to HIV prevalence among men who have sex with men and in the general population. Goodness of fit was estimated from the results. The target population was that between 15 and 49 years and sexually active. We compared a strategy of early antiretroviral treatment delivery (CD4+> 350cell/mL) for all infected individuals versus a strategy targeted to men who have sex with men. The impact at population level for both strategies was estimated as the HIV incidence reduction for the period between 2013 and 2030 or 2040.