Workshop participants discussed: the role of HIV subtypes in disease; the treatment of oral candidiasis; the relationship between and among viral load, CD4+ counts, oral candidiasis and oral hairy leukoplakia, pigmentation; and the development of a reliable oral index to predict disease progression. Regarding HIV, the literature revealed that Type I (HIV-I), in particular group M, is involved in the majority (90%) of documented infections, and groups N and O to a lesser extent. Viral envelope diversity led to the subclassification of the virus into nine subtypes, or clades-A-D, F-H, J, and K-each dominating in different geographical areas. HIV-2, currently occurring mostly in West Africa, appears to be less virulent. No evidence could be produced of any direct impact of type, subtype, or clade on oral lesions, and participants believed that further research is not feasible. Oral candidiasis in patients from resource-poor countries should be prevented. When the condition does occur, it should be treated until all clinical symptoms disappear. Oral rinsing with an antimicrobial agent was suggested to prevent recurrence of the condition, to reduce cost, and to prevent the development of antifungal resistance. Lawsone methyl ether, isolated from a plant (Rhinacanthus nasutus leaves) in Thailand, is a cost-effective mouthrinse with potent antifungal activity. Evidence from a carefully designed prospective longitudinal study on a Mexican cohort of HIV/AIDS patients, not receiving anti-retroviral treatment, revealed that the onset of oral candidiasis and oral hairy leukoplakia was heralded by a sustained reduction of CD4+, with an associated sharp increase in viral load. Analysis of the data obtained from a large cohort of HIV/AIDS patients in India could not establish a systemic or local cause of oral melanin pigmentation. A possible explanation was a dysfunctional immune system that increased melanin production. However, longitudinal studies may contribute to a better understanding of this phenomenon. Finally, a development plan was presented that could provide a reliable prediction of disease progression. To be useful in developing countries, the index should be independent of costly blood counts and viral load.