BACKGROUND: Given the great efforts put into the strategic objective of reducing the proportion of HIV-infected patients that are undiagnosed, the aim of the present study was to review the temporal trends between 1997 and 2016 for median estimates of infection duration and median CD4 count at diagnosis for the main patient origins in French Guiana. METHODS: CD4 cell count at HIV sero-conversion and square root of CD4 cell decline were obtained using the CD4 decline in a cohort of HIV-infected persons in the UK, fitting random effect (slope and intercept) multilevel linear regression models. Multivariate analysis used robust regression for modeling the delay between estimated HIV seroconversion and diagnosis and quantile regression for CD4 at HIV diagnosis. RESULTS: The median interval between the estimated HIV seroconversion and HIV diagnosis was 8 years for patients fromBrazil, 4.5 years for those from Haiti, 6.6 years for those from Suriname, 3.3 years for patients from Guyana, and 3.1 years for French patients. A simple robust regression model with French patients as reference group adjusting for sex and age at the time of diagnosis showed that the interval was significantly longer for Brazilian (β = +3.7 years, P = 0.001), Surinamese (β = +4.2 years, P<0.0001), Haitian origins (β = +1.5 years, P = 0.049) but not for those originating from Guyana (β = -0.03 years, P = 0.9); Men independently had a longer interval than women (β = +3.5 years, P<0.0001). CONCLUSIONS: Despite great efforts in French Guiana regarding HIV testing both in terms of diversification and intensification we still need to tailor the offer to better reach the communities in need. These results should help authorities scale up and optimize initiatives to reduce the proportion of patients who are unaware of their infection. They also raise the question of the role of stigma and discrimination as a barrier to HIV testing in small communities, and further emphasize the importance of reducing it