Background: We report the final results of Community Programs for Clinical Research on AIDS (CPCRA-064) study, a multicenter, prospective, randomized, controlled trial that determines the long-term clinical impact of structured treatment interruption (STI) in patients with multidrug-resistant (MDR) HIV-1., Methods and Results: Two hundred seventy-four patients on stable antiretroviral therapy with MDR HIV-1 treatment failure were randomized to a 4-month STI, followed by an optimized antiretroviral regimen (STI arm, n = 140) or an immediate change to an optimized antiretroviral regimen (control arm, n = 134). Main outcome measures were progression of disease or death and changes from baseline in HIV RNA levels (log copies/mL) and CD4 cell counts (cells/mm). The median baseline HIV RNA level was 5.0 log copies/mL, the median CD4 count was 147 cells/mm, and the nadir CD4 count was 32 cells/mm. The median follow-up was 37 months. After the STI period, there were no differences in HIV RNA level responses between treatment arms. Differences in CD4 count responses always favored the control arm, with an advantage of 84 cells from 0 to 4 months (P < 0.0001), 50 cells from 4 to 12 months (P < 0.0001), 45 cells from 12 to 24 months (P = 0.006), and 43 cells after 24 months (P = 0.07). Rates in the STI and control arms for first progression-of-disease event or death were 17.5 and 14.3, respectively (hazard ratio = 1.28; P = 0.22)., Conclusion: STI before changing regimens in patients with MDR HIV-1 treatment failure has a prolonged negative impact on CD4 cell count recovery and does not confer progression of disease or virologic benefits.