tRNA(3Lys) is a primer for reverse transcription in human immunodeficiency virus type 1 (HIV-1), and the anticodon of tRNA(3Lys) has been implicated in playing a role in both its placement onto the HIV-1 genome and its interaction with HIV-1 reverse transcriptase (RT). In this work, the anticodon in a tRNA(3Lys) gene was changed from UUU to CUA (tRNA(3Lys)Su+) or, in addition, G-73 was altered to A (tRNA(3Lys)Su+G73A). COS-7 cells were transfected with either wild-type or mutant tRNA(3Lys) genes, and both the wild-type and mutant tRNA(3Lys) produced were purified by using immobilized tRNA-specific hybridization probes. Each mutant tRNA(3Lys) was tested for its ability to prime reverse transcription in vitro, either alone or in competition with wild-type tRNA(3Lys). Short RT extensions of wild-type and mutant tRNALys could be distinguished from each other by their different mobilities in one-dimensional single-stranded conformation polymorphism polyacrylamide gel electrophoresis. These reverse transcription products show that heat-annealed tRNA(3Lys)Su+ has the same ability as heat-annealed wild-type tRNA(3Lys) to prime RT and competes equally well with wild-type tRNA(3Lys) for priming RT. tRNA(3Lys)Su+G73A has 60% of the wild-type ability to prime RT but competes poorly with wild-type tRNA(3Lys) for priming RT. However, the priming abilities of wild-type and mutant tRNA(3) are quite different when in vivo-placed tRNA is examined. HIV-1 produced in COS cells transfected with a plasmid containing both the HIV-1 proviral DNA and DNA coding for tRNA(3Lys)Su+ contains both endogenous, cellular wild-type tRNA(3Lys) and mutant tRNA(3Lys). When total viral RNA is used as the source of primer tRNA placed onto the genomic RNA in vivo, only wild-type tRNA(3Lys) is used as a primer. If the total viral RNA is first heated and exposed to hybridizing conditions, then both the wild-type and mutant tRNA(3Lys) act as primers for RT. These results indicate that the tRNA(3Lys)Su+ packaged into the virions is unable to act as a primer for RT, and a model is proposed to explain the disparate results between heat-annealed and in vivo-placed primer tRNA.