BACKGROUND: Episodic use of antiretroviral therapy guided by CD4+ cell counts is inferior to continuous antiretroviral therapy. OBJECTIVE: To determine whether reinitiating continuous antiretroviral therapy in patients who received episodic treatment reduces excess risk for opportunistic disease or death. DESIGN: Randomized, controlled trial. SETTING: Sites in 33 countries. PATIENTS: 5472 HIV-infected individuals with CD4(+) cell counts greater than 0.350 x 10(9) cells/L enrolled from January 2002 to January 2006. INTERVENTION: Episodic or continuous antiretroviral therapy initially, followed by continuous therapy in participants previously assigned to episodic treatment. MEASUREMENTS: Opportunistic disease or death was the primary outcome. RESULTS: Eighteen months after the recommendation to reinitiate continuous therapy, mean CD4+ cell counts were 0.152 x 10(9) cells/L (95% CI, 0.136 to 0.167 x 10(9) cells/L) less in participants previously assigned to episodic treatment (P < 0.001). The proportion of follow-up time spent with CD4+ cell counts of 0.500 x 10(9) cells/L or more and HIV RNA levels of 400 copies/mL or less was 29% for participants initially assigned to episodic therapy and 66% for those assigned to continuous therapy. Participants who reinitiated continuous therapy experienced rapid suppression of HIV RNA levels (89.7% with HIV RNA levels ≤ 400 copies/mL after 6 months), but CD4+ cell counts after 6 months remained 0.140 x 10(9) cells/L below baseline. The hazard ratio (episodic versus continuous treatment) for opportunistic disease or death decreased after the recommendation to reinitiate continuous therapy (from 2.5 [CI, 1.8 to 3.5] to 1.4 [CI, 1.0 to 2.0]; P = 0.033 for difference). The residual excess risk was attributable to failure to reinitiate therapy by some participants and slow recovery of CD4+ cell counts for those who reinitiated therapy. LIMITATION: Follow-up was too short to assess the full effect of switching from episodic to continuous antiretroviral therapy. CONCLUSION: Reinitiating continuous antiretroviral therapy in patients previously assigned to episodic treatment reduced excess risk for opportunistic disease or death, but excess risk remained. Episodic antiretroviral therapy, as used in the SMART study, should be avoided.