To measure the extent of HIV resistance to (-)-2',3'-dideoxy-3'-thiacytidine (3TC, lamivudine) within the context of monotherapy and to assess the presence of the M184V substitution in the case of 3TC-resistant viruses. Whether the success of 3TC in clinical trials could be due, in part, to an increase in the fidelity of HIV reverse transcriptase conferred by the M184V substitution was also considered. Two separate monotherapy studies were evaluated, one involving adults with CD4 counts > or = 300 x 10(6)/l, and the second involving children, some of whom had received antiretroviral treatment previously, while others were drug naive. Peripheral blood and plasma samples were collected regularly, and HIV isolation and determinations of drug median inhibitory concentration values were performed using umbilical cord mononuclear cells as targets. Amplification of the 184 mutation was performed by the polymerase chain reaction, using specific primer pairs. Fidelity determinations using purified, recombinant HIV reverse transcriptase derived from either wild-type virus or viruses that contained the 184V substitution were performed. Phenotypic resistance was detected in almost all subjects at times ranging from 8-20 weeks after initiation of therapy. The 184V substitution was usually detected prior to the occurrence of phenotypic resistance to 3TC. Fidelity determinations revealed that the 184V substitution conferred an approximately 5- to 10-fold increase in HIV reverse transcriptase fidelity. In addition, titres of patient sera tested for their ability to neutralize autologous sequential viral isolates were stabilized in patients receiving 3TC therapy as opposed to other drugs. Resistance to 3TC developed in virtually all subjects treated with this drug, and was associated with the appearance of an M184V mutation in HIV reverse transcriptase. The clinical benefit of 3TC therapy may be attributable in part to selection of viruses that are less able to replicate and mutate than the wild types.