Background: High-income countries studies show unfavorable trends in stroke incidence (SI) in younger populations. We aimed to estimate temporal change in SI disaggregated by age and sex in Latin America and the Caribbean region (LAC). Methods: A search strategy was used in MEDLINE, WOS, and LILACS databases from 1997 to 2021, including prospective observational studies with age and sex-disaggregated data of first-ever stroke (FES) incidence. Risk of bias was assessed with The Joanna Briggs Institute's guide. The main outcomes were incidence rate ratio (IRR) and relative temporal trend ratio (RTTR) of SI, comparing time periods {greater than or equal to}2010 with <2010. Pooled RTTR (pRTTR) only considering studies with two periods in the same population were calculated by random-effects meta-analysis. Results: From 9,242 records identified, six studies were selected including 4,483 FES in 4,101,084 individuals. Crude IRR in younger subjects (<55 years) comparing {greater than or equal to}2010:<2010 periods showed an increase in SI in the last decade (IRR 1.37;95%CI 1.23-1.50), in contrast to a decrease in older people during the same period (IRR 0.83; 95%CI 0.76-0.89). Overall RTTR (<55:{greater than or equal to}55 years) was 1.65 (95CI% 1.50-1.80), with higher increase in young women (pRTTR 3.08; 95%CI 1.18-4.97; p for heterogeneity <0.001). Conclusions: An unfavorable change in SI in young people - especially in women - was detected in the last decade in LAC. Further investigation of the explanatory variables is required to ameliorate stroke prevention and inform local decision-makers. Registration of protocol: CRD42022332563 (PROSPERO).