To determine the extent to which genotype (GT) or phenotype (PT) methods provide HIV-1 drug resistance information that is overlapping or complementary, both tests were performed on 1378 patient plasma samples. Discordance, defined as determination of reduced susceptibility measured by PT but sensitivity by GT (PT-R/GT-S), or vice versa (PT-S/GT-R), was common: 83, 62, 43, and 28% of samples with evidence of drug resistance had at least 1, 2, 3, or 4 drugs discordant, respectively. Three types of discordance were observed: PT-R/GT-S, and PT-S/GT-R with or without the presence of mixtures at resistance-associated positions (25%, 34%, and 41% of all discordance, respectively). After accounting for mixtures, results for didanosine (30%), zalcitabine (18%), tenofovir (17%), abacavir (14%), lamivudine (12%), and amprenavir (11%) were discordant in >or= 10% of samples. PT-S/GT-R results were most common for didanosine and zalcitabine, whereas PT-R/GT-S results were most common for lamivudine and amprenavir. PT provided quantitative assessment of the degree of reduced susceptibility and identified reduced susceptibility (PT-R/GT-S) or normal susceptibility (PT-S/GT-R) that was not recognized by the GT interpretation algorithm. GT provided valuable information when mixtures were present and minor populations of drug resistant virus were not detected by phenotyping (PT-S/GT-R results). This demonstrates the complementary nature of information provided by PT and GT tests and suggests that their combined use can provide additional clinically-relevant information.