Objective: To investigate the performance of the HIV RDTs used in Zambia. Method: 2,564 participants aged between 15 and 95 years from two sites in Lusaka province years were tested on OraQuick ADVANCE, Abbot Determine, and then confirmed on Uni-Gold Recombigen. The data from the participants were analyzed using SPSS version 25.0. Results: The 3 RDTs when compared to the 4th generation Abbot Architect results had the following results: OraQuick ADVANCE, Alere Determine and Uni-Gold Ultra, at 95% CI had Sensitivities of: 91.8%, 93.3% and 92.5% respectively. The specificities of OraQuick ADVANCE and Uni-Gold were the same (100.0%; 95% CI: 98.8 -100.0) but slightly different from Alere Determine (99.8%). Positive predictive values at 95% CI were 100% for OraQuick ADVANCE and Uni-Gold and 98.4% for Alere Determine. Negative predictive values (at 95% CIs) were 99.1, 99.2 and 99.1 for OraQuick ADVANCE, Alere Determine, and Uni-Gold Ultra respectively. The results showed that these RDTs could only detect 12 out of every 13 HIV positive results. Conclusion: Third generation RDTs are not effective in detecting acute positive cases. Fourth generation Rapid Tests are required to capture the positive cases being missed out.