Data on genotypic and phenotypic resistance testing of HIV-1 in the routine clinical setting are lacking. In a retrospective single-center study, all patients (n = 102) for whom genotypic resistance typing (GRT) and phenotypic resistance typing (PRT) were performed during the calendar year 2002 were examined. GRT and PRT results were concordant for 79% of the drugs, being highest for nevirapine (92%) and lowest for didanosine (57%). Concordance of results for protease inhibitors was lowest for lopinavir (78%) and highest for indinavir (88%). Discordant results for lamivudine were observed in 16% of patients; 90% of these results corresponded to high-level resistance by PRT and susceptibility by GRT. Overall, HIV loads were lower and CD4+ cell counts higher after therapy following resistance testing, but a significant association with the number of active drugs as predicted by GRT or PRT could not be identified. In a subgroup of 43 patients with virological failure under antiretroviral therapy and sufficient follow-up data, HIV loads were significantly lower after 3 and 6 months. More patients with HIV loads <400/ml had 2 or more active drugs according to PRT (21/29 [75%]) than according to GRT ([15/29 [52%]; p = 0.109. This was also found for HIV loads <50/ml (PRT 16/22 [72%], GRT 10/22 [42%]; p = 0.103), although the differences were not statistically significant. There was no discernable difference between GRT and PRT in the clinic-based population, but the numbers of resistance tests performed are not sufficient to draw definitive conclusions.