<h3>Introduction</h3> Guatemala has a concentrated HIV epidemic among key populations; in 2003, the first rapid test validation in Central America was performed. A national HIV testing algorithm with a rapid test and a second test using an enzyme immunoassay (EIA) was developed (average turnaround time of 2 weeks). Given the changes in the market since 2003, we present results of a second rapid test validation to update the Guatemalan algorithm. <h3>Methods</h3> Validation was performed in 2012–2013, evaluating a total of 10 rapid tests in two phases: 1) with serum, HIV-negative samples were obtained from blood banks, and positive samples from HIV care units; 2) with whole blood, negative samples were obtained from antenatal care services and positive samples from HIV care units. We tested 452 serum samples [244 HIV+, 208 HIV-] and 432 whole blood samples (10 HIV positive, 422 HIV negative) using two EIA (Roche ELECSYS HIV Combi and Abbott AXSYM HIV Ag/Ab Combo). Discordant results were evaluated with HIV-1 Western Blot. For fourth generation rapid tests, only antibody was evaluated. Data was analysed using OpenEpi. <h3>Results</h3> Six HIV rapid tests were included for both phases. All tests (except Double Check Gold) reported sensitivity higher than 99% and specificity higher than 95%. Determine HIV 1/2 and HIV Ag/Ab reported the highest results for sensitivity (100%). Hexagon HIV (100%) and Accu-Tell (99.5%) reported the highest results for specificity. All rapid tests that also included HIV-2 detection, showed cross-reactivity, ranging from 37% of HIV cases for Anarapid HIV 1/2/O Tri-line to 15% for Rapid HIV 1/2/O Tri-Line. <h3>Conclusion</h3> Our results show that Determine HIV½ and HIV Ag/Ab should be recommended as a screening test while Hexagon HIV and Accu-Tell should be recommended as second test. Rapid test validation provides valuable information for on-site confirmation of reactive results improving diagnosis turnaround time. <h3>Disclosure of interest statement</h3> We declare that we have no conflicts of interest.