The main objectives of the work presented in this thesis were to develop an assay to subtype viruses circulating in Uganda and to study virological and immunological factors that might be important in HIV-1 vaccine development in East Africa. Development of an antibody binding assay for serological subtyping revealed a complex pattern of cross-reactivity, and led to further investigation of these responses. Competitive inhibition and adsorption studies identified a spectrum of antibodies in the human antisera, possibly due to multiple immune responses to more than one viral epitope or to one immunodominant epitope. Neutralizing antibodies to specific V3 loop peptides were raised in rabbits; MN and U31 peptides were good immunogens, but Ugandan consensus sequence peptides raised antisera with weak peptide reactivities. Antibody binding assay of partial V3 loop peptides indicated that most Ugandan antisera recognised an epitope(s) in the C-terminal region of the V3 loop, not at the apex. Virus samples from the UVRI clinic and the Natural History cohort in rural SW Uganda were isolated and characterized. Comparison of primary isolates in PBMC cultures and cell adapted variants indicated that SI and NSI variants may be present within a primary isolation culture. Neutralizing antibody titres were compared between cell line adapted viruses and primary isolates: Differences found led to the use of low passaged primary isolates for this investigation. Cross-clade neutralizing antibodies were present in most antisera, many with high neutralization titres (>640); suggesting a shared immunogenic epitope, which could be important in vaccine design. Analysis of sequential serum samples revealed no decrease in neutralizing titres with time. One sequential virus isolate (3029), re-isolated after 2 years, was more resistant to neutralization by heterologous antisera. No correlation was established between antibody binding and neutralization by human or rabbit antisera. Comparison of clinical data with neutralization data was inconclusive.