To determine the long-term immunologic and virologic effects of antiretroviral treatment interruptions, a retrospective analysis of an ongoing observational database was performed at a university HIV clinic. All patients who began highly active antiretroviral therapy (HAART) after January 1, 1996 and (1) were HAART experienced for >/=90 days, (2) had a treatment interruption (TI) for >/=30 days, (3) resumed HAART for >/=30 days, and (4) had CD4(+) cell counts performed pre- and post-TI were included. Main outcome measures included the following: Immunologic success was defined as a post-TI CD4(+) cell count >90% of the pre-TI CD4(+) cell count (post-TI/pre-TI, >90%). Virologic success was defined as a post-TI viral load (VL) less or equal to twice the pre-TI VL (post-TI/pre-TI, </=2) or a post-TI VL of <1000 copies/ml. The pre-TI (baseline) value was the value at the start of the TI (range, -20 to +7 days); the post-TI value was the highest CD4(+) cell count and lowest VL copy achieved during the follow-up window (270 days). One thousand and eight patients were included in the analysis and 75 met the inclusion criteria. Forty-four of 75 patients (58.6%) achieved a successful immunologic outcome and 52 of 68 patients (76.5%; 7 patients did not have a VL determined within the specified periods) achieved a successful virologic outcome. No factors predicting success were identified. The median CD4(+) cell counts pre- and post-TI were 233 and 231 cells/microl, respectively; the median VLs pre- and post-TI were 11,456 and 404 copies/ml, respectively. We conclude that the majority of our patients in virologic failure who underwent a temporary TI recovered 90% of their baseline CD4(+) cell counts and returned to within 2-fold of their baseline VL when HAART was resumed.