Rapid tests allow outpatient, low cost, reliable, screening for chronic HIV infection. However, data regarding their sensitivity on primary infection remain scarce. The objective of this study was to assess sensitivity of nine HIV rapid tests for primary HIV-1 infection screening. Seventy-five serum samples from patients during HIV-1 primary infection were included. Primary infection was diagnosed by a positive 4th generation ELISA and HIV-1 RNA positivity confirmed by Western blot patterns associated with HIV-1 primary infection. Early seroconversion was defined as the absence of antibodies on HIV-1 Western blot associated with HIV-1 RNA and p24-antigen positivity. An identical sensitivity (95% CI) of 76.7% (65.2-84.2%) was observed for HIV 1/2 STAT-PAK Assay (STAT-PAK), INSTI™ HIV-1/HIV-2 antibody Test (INSTI), SURE CHECK HIV 1/2 (SURE CHECK) and MULTISURE HIV rapid test (MULTISURE) with visual reading. Sensitivity was 74.7% (63.8-83.1%) for MULTISURE (automatic reading), 77.0% (66.3-85.1%) for FIRST RESPONSE Test VIH 1-2.O CARTE (FIRST RESPONSE), 83.8% (73.8-90.5%) for VIKIA HIV1/2 (VIKIA), 88.0% (78.7-93.6%) for Genie™ Fast HIV 1/2 (Genie Fast), 88.6% (79.0-94.1%) for Hexagon HIV (Hexagon), and 92.8% (83.6-96.3%) for Exacto TEST HIV Pro (Exacto). However, rapid tests performed poorly for the early seroconversion subgroup (n = 14), with sensitivities ranging from 7% (1.3-31.5%) for STAT-PAK, INSTI, SURE CHECK, MULTISURE (automatic reading), to 29% (12-55%) for FIRST RESPONSE, 31% (13-58%) for VIKIA, 43% (21-67%) for Hexagon and 57.1% (32.6-78.6%) for Exacto and Genie Fast. Overall, despite significant discrepancies in sensitivity, HIV rapid tests should be used with caution in the context of a suspected primary infection.