The purpose of this study is to model the potential economic impact of viral load-driven triple drug combination (including a protease inhibitor) antiretroviral therapy on incremental drug and hospitalisation costs among individuals with HIV disease. Individuals included in the study were HIV-positive men and women from the province of British Columbia, Canada, who were aged 18 years or older and had given consent to access their medical records. The study employed pharmacoeconomic modelling of drug- and hospital-utilisation patterns among a population-based cohort with free access to antiretroviral therapy. Protease inhibitor use and associated costs based on actual use in a subsequent period was modelled upon men and women who were able to maintain stable CD4+ cell counts (slope > or = 0) for at least 6 months (baseline period) with an average follow-up period of 30 months (protease-like group). A control was modelled upon individuals with declining CD4+ cell counts (slope < 0) during similar baseline and follow-up periods. The primary outcome measure was average annual incremental cost of triple drug therapy net of hospitalisation and testing costs in 1996 Canadian dollars ($Can). The utilisation pattern of drugs and hospitals was modelled from actual use among a total of 1271 individuals who were eligible for this analysis. Programme participants who gave consent to access their medical records were more likely to be men (p < 0.001), older (p < 0.020), and on antiretroviral therapy (p < 0.001) than programme participants who did not give consent. No differences were observed between the protease-like and comparison groups with respect to age (p = 0.65) and CD4+ cell count (p = 0.30) at study entry. Over a period of 1 year, the protease-like group was shown to spend less time in hospital (2.7 vs 6.6 days; p < 0.001). This difference in hospitalisation remained in multivariate models, adjusting for prior AIDS-defining illnesses and gender. The average annual incremental cost of adding a protease inhibitor to a 2-drug antiretroviral regimen was estimated to be $Can2318 per person. The cost implications of hospital stay while using a protease inhibitor drug and 2 nucleosides translated into an average annual incremental cost (savings if negative) of between -$Can4798 and -$Can2227 per person. The overall average annual incremental cost impact per person associated with triple drug therapy with a protease inhibitor varied between -$Can2288 to $Can283. Negative incremental costs imply overall savings from adopting triple combination therapy. This modelling exercise demonstrated that the cost of triple drug combination antiretroviral therapy with a protease inhibitor among HIV-positive men and women was considerably less than the expected acquisition cost of the drug alone due to hospitalisation savings in the province of British Columbia.