The purpose of this study was to analyze the agreement between self-reported and device-based sedentary time among eight countries in Latin America. As part of the Latin American Study of Nutrition and Health (ELANS), data were collected from 2524 participants (18–65 years) across eight countries. Participants reported time spent sedentary in different activities (computer use at home, videogame use, reading, sitting down to chat with friends/relatives or listening to music, speaking on the phone, watching TV, and riding in a car). Overall sitting time was assessed using a single item from the International Physical Activity Questionnaire (IPAQ). Device-based sedentary time was assessed using Actigraph GT3X accelerometers. Self-reported overall sitting time (227.1 min/day) produced the lowest values of the three assessment methods, followed by self-reported sum of different types of sedentary behavior (364.1 min/day) and device-based sedentary time (568.6 min/day). Overall, correlation coefficients and ICC varied from weak to moderate (rho: 0.25–0.39; ICC: 0.21:0.39) between self-reported sum of different types of sedentary behavior, self-reported overall sitting time, and device-based sedentary time. The Bland-Altman plots indicated low to moderate agreement between self-reported overall sitting time and device-based sedentary time by sex. Self-report measures underestimate sedentary behavior and overall sitting time when compared with device-based measures. The weak and moderate level of agreement between methods indicates that caution is required when comparing associations between different self-report and device-based measures of sedentary behavior with health outcomes.