BACKGROUND: Antibody responses as serological markers of Plasmodium vivax infection have been shown to correlate with exposure, but little is known about the other factors that affect antibody responses in naturally infected people from endemic settings. To address this question, we studied IgG responses to novel serological exposure markers (SEMs) of P. vivax in three settings with different transmission intensity. METHODOLOGY: We validated a panel of 34 SEMs in a Peruvian cohort with up to three years' longitudinal follow-up using a multiplex platform and compared results to data from cohorts in Thailand and Brazil. Linear regression models were used to characterize the association between antibody responses and age, the number of detected blood-stage infections during follow-up, and time since previous infection. Receiver Operating Characteristic (ROC) analysis was used to test the performance of SEMs to identify P. vivax infections in the previous 9 months. PRINCIPAL FINDINGS: Antibody titers were associated with age, the number of blood-stage infections, and time since previous P. vivax infection in all three study sites. The association between antibody titers and time since previous P. vivax infection was stronger in the low transmission settings of Thailand and Brazil compared to the higher transmission setting in Peru. Of the SEMs tested, antibody responses to RBP2b had the highest performance for classifying recent exposure in all sites, with area under the ROC curve (AUC) = 0.83 in Thailand, AUC = 0.79 in Brazil, and AUC = 0.68 in Peru. CONCLUSIONS: In low transmission settings, P. vivax SEMs can accurately identify individuals with recent blood-stage infections. In higher transmission settings, the accuracy of this approach diminishes substantially. We recommend using P. vivax SEMs in low transmission settings pursuing malaria elimination, but they are likely to be less effective in high transmission settings focused on malaria control.