The identification of HIV strains that are resistant to antiretroviral drugs, which emerge during a patient's therapy or are already present in infected individuals prior to treatment, is of increasing importance for the clinical management of HIV-infected persons. Two different methods were compared for the genotypic detection of resistance development in the HIV-1 reverse transcriptase (RT) gene, the commonly used sequence analysis, and the commercially available RT-line immunoprobe assay (LIPA), which can detect mutations at six separate codons of the RT gene, which are known to confer resistance to certain nucleoside inhibitors. Eighty serum samples from HIV-1-infected persons, some of whom were receiving antiretroviral therapy, were investigated in parallel by sequencing as well as by LIPA. LIPA results agreed with sequence data in the vast majority of the cases. However, in 40% of the samples, LIPA failed to yield evaluable results for one or more of the codon positions. In particular, LIPA detection rate was low at codon 41 (75%), whereas at codons 69/70, 74, 184, and 215 results were obtained from 90%-95% of the samples. A number of mutations in the vicinity of the respective codons were detected by sequencing, and these may have been responsible for the LIPA hybridization failure. There remained a number of samples, however, where no explanation for the lack of hybridization could be derived from sequence data. Our results indicate that the use of the LIPA does not eliminate the need for sequence analysis for detection of drug-resistant HIV strains.