Background: Although the number of HIV infected people receiving highly active anti-retroviral therapy (HAART) in low- and middle- income countries has increased dramatically, optimal disease management in these settings is not well defined. Methods: We developed a model to compare the lifetime costs and benefits of three different strategies for managing HAART: a symptom-based strategy, a CD4-based strategy, and a CD4 & viral load strategy for starting, switching, and stopping HAART. Patients were followed from the time of presentation to care, and were eligible for first- and second-line regimens. We used published data from Cape Town cohorts and performed a cost-effectiveness analysis. Results: Compared to a symptom-based approach, monitoring CD4 every 6 months and starting treatment at a threshold of 200 cells/μl was associated with a gain of 10.8 months of life (57.6 vs. 68.4 discounted months) and lifetime cost savings of $689 per person ($3,752 vs. $4,441). CD4-based strategies where treatment was started at 350 cells/μl were more costly but yielded an additional gain of 5.7 months compared to a threshold of 200 cells/μl. Using viral load in addition to CD4 was substantially more expensive than monitoring CD4 alone, produced a gain of 1.4 months of life, and had an incremental cost-effectiveness ratio of $6,050 per life-year gained. In sensitivity analysis, the price of viral load testing and the rates of virologic failure had a significant effect on the cost-effectiveness of viral load monitoring. Conclusion: Expanding access to CD4 monitoring in resource-limited regions is cost-saving and provides substantial health benefits. The cost-effectiveness of viral load monitoring depends on the price per test and rates of virologic failure.