Background: HIV-2 is less virulent than HIV-1 but can cause AIDS indistinguishable from AIDS caused by HIV-1. Guinea-Bissau is believed to be the nucleus of the HIV-2 epidemic and the prevalence was >20% in the 1980s in elderly in Caio, a rural area. In 1990-2007, HIV-2 prevalence halved and in 2007 infection amongst young adults in the region was very low (<1% in 15-34 year olds). However, HIV-2 prevalence was still 12% amongst individuals aged ≥45, due in part to increased susceptibility of women to HIV-2 with increasing age. This suggests that HIV-2 may persist as an infection of the elderly. Through a model we aimed to answer the following questions: will HIV-2 in Caio go extinct and if so, when would this happen, or will it persist as an infection among elderly? Methodology: A deterministic transmission model was developed and fit to the age-stratified incidence and prevalence data from the HIV-2 epidemic in Caio, as measured in 3 serosurveys in 1990, 1997 and 2007. In the model the population is segregated into 3 states: Susceptible, Infected and Removed (SIR model). A stochastic version of the model was also formulated by allowing death and infection to be governed by probabilistic methods and this version was used with 1000 simulations to make future predictions of HIV-2 prevalence and incidence. Results: Our model predicts that HIV-2 infections (including dual HIV-1/2 infections) will continue to decline in prevalence and become extinct in Caio this century. HIV-2 prevalence is expected to fall below 0.1% by 2053 (95% CI: 2043-2074) (Figure 1A). New infections are predicted to cease around 2047 (2029-2083) (Figure 1B) but complete extinction is expected to take longer. Mortality rates are relatively low for HIV-2, as compared to HIV-1, meaning that the last few individuals to acquire infection may survive for decades. Extinction is expected to occur around 2071 (2054-2106). Conclusions: Based on our model, it is predicted that HIV-2 infection will continue to decline rapidly in Caio such that new infections will cease and prevalence will reach low levels. HIV-2 will go extinct in Caio during the second half of this century. HIV-2 epidemics in other parts of the world have behaved in a very similar way thus far, making global extinction of HIV-2 a possibility. As low plasma viral loads play a key role in the restricted transmission and therefore decline of HIV-2, we highlight the relevance of our results to the potential control of the HIV-1 epidemic via universal treatment.