PURPOSE OF REVIEW: To discuss the implications of the results of the Strategies for Management of Antiretroviral Therapy Study and other treatment interruption studies on the question of when antiretroviral therapy should be initiated. RECENT FINDINGS: In the Strategies for Management of Antiretroviral Therapy Study, CD4 count-guided, episodic use of antiretroviral therapy as compared with continuous antiretroviral therapy resulted in an increased risk of all-cause mortality, almost entirely due to causes other than AIDS, and a composite outcome of cardiovascular disease, renal disease and liver disease. Subgroup analyses in the Strategies for Management of Antiretroviral Therapy Study indicated that the increased risk in the episodic antiretroviral therapy group compared with the continuous antiretroviral therapy group was evident in patients taking antiretroviral therapy at entry (antiretroviral therapy stopped after randomization) and in patients not taking antiretroviral therapy at entry (patients remained off antiretroviral therapy until their CD4 count declined to below 250 cells/mm). SUMMARY: The Strategies for Management of Antiretroviral Therapy Study did not directly address the 'when to start' question. Data from the Strategies for Management of Antiretroviral Therapy Study and other studies strongly suggest, however, that the use of antiretroviral therapy earlier than recommended by current guidelines warrants investigation. Definitive data from randomized studies that are powered to reliably assess risks and benefits are needed to guide when antiretroviral therapy is initiated.