Objectives To project the clinical and economic outcomes of a genotype assay for selection of third-line antiretroviral therapy (ART) in resource-limited settings, as per the planned international A5288 trial (MULTI-OCTAVE). Methods We used the Cost-effectiveness of Preventing AIDS Complications (CEPAC)-International Model to compare three strategies for subjects who have failed second-line ART in South Africa: (1) Sustained second-line: no genotype assay, all subjects remain on second-line ART; (2) A5288: genotype to determine the resistance profile and assign an appropriate regimen; or (3) Population-based third-line: no genotype, all subjects switch to a potent third-line regimen. Model inputs are from published data in South Africa. Resistance profiles, ART regimens, and efficacy data were those used for trial planning. Results Projected life expectancy for sustained second-line, A5288, and population-based third-line are 61.1, 103.8, and 104.2 months. Compared to sustained second-line ($12,460), per person lifetime costs increase for the A5288 ($39,250) and population-based ($44,120) strategies. The incremental cost-effectiveness ratio of A5288, compared to sustained second-line, is $7,500/year of life saved (YLS), and for population-based third-line, compared to A5288, is $154,500/YLS. In the A5288 strategy, very late presentation to care, coupled with lengthy delays to obtain the genotype, dramatically reduces 5-yr survival, making the population-based third-line strategy more attractive. Conclusions We project that, while the public health approach to third-line therapy is unaffordable, genotype assays and third-line ART in resource-limited settings will increase survival and be cost-effective compared to the population-based approach, supporting the value of an efficacy study