Since 1987 major advances have been made in our understanding of the pathogenesis of infection and the possible inhibition of the HIV virus. Various drugs targeted to the different steps of viral replication have been selected, but drugs such as soluble CD4 or dextran derivatives aimed to inhibit or interfere with the GP120-CD4 attachment step have shown little or no clinical benefit. Protease inhibitors or interferons acting at the post-transcriptional level are currently under phase I to II investigation. The only group of compounds clinically active belong to the nucleosides analogues that act as DNA chain terminators and by inhibiting viral reverse transcriptase. Zidovudine (AZT), didanosine (ddl) and dideoxycytidine (ddC) have been extensively studied, and used on a large clinical scale. Stavudine (D4T), deoxyfluorothymidine (FLT) and 3′thiacytidine (3TC) are entering phase I–II studies. Being the first nucleoside analogue discovered and used since early 1985, zidovudine remains the gold standard of anti-HIV therapy. Zidovudine is indicated at a dosage varying between 500 and 1000 mg in patients with AIDS and ARC, in asymptomatic patients with CD4 < 200/mm3 and in asymptomatic patients with CD4 between 200 and 500 cells/mm3 with a rapid decrease of CD4 cell count or a positive P24 circulating antigen. There is as yet no consensus concerning patients with more than 500 cell/mm3. ddl and ddC are currently indicated for patients intolerant to AZT or in those who have not responded (clinically or biologically) to AZT. Emergence of resistance to AZT has been reported in 30 to 80% of patients at various stages of the disease and after six to nine months of therapy. There are indications that this resistance is clinically relevant, and may show a need for early combination therapy. Trials are now starting or under way comparing AZT alone with various combinations of ddl and ddC. The best combination therapy should involve drugs acting at different sites of viral replication such as nucleosides analogues and protease inhibitors. The group of specific HIV1 reverse transcriptase inhibitors has been disappointing. Poor bioavailability or very rapid emergence (after three weeks) of resistance have precluded the clinical use of TIBO-like or TIBO derivatives which were highly active in vitro.