INTRODUCTION: In Brazil, the prevalence and costs of pain will increase substantially with population ageing. Understanding of pain epidemiology is needed for the development of health care policies that can minimize this projected burden. OBJECTIVE: To investigate the prevalence of pain and associated factors at baseline of the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). METHODS: Data were collected in public institutions of higher education/research (2008-2010). Pain in the past 30 days and pain attributed to psychological distress ("with psychological attributions"-PPA) were evaluated by the Clinical Interview Schedule-Revised (CIS-R). The independent t-test and χ2 test investigated associations between sociodemographic/clinical factors and each pain episode. Multivariable analyses including age, sex, leisure-time physical activity, depression, and arthritis/rheumatism, and factors showing univariate associations at the P < 0.10 level, were performed. RESULTS: Fifteen thousand ninety-five civil servants were included (52.1 ± 9.1 years, 54.4% female). The prevalence of any pain was 62.4% (95% confidence interval 61.6%-63.2%), and of PPA was 22.8% (95% confidence interval 22.2%-23.5%). Factors associated with any pain and PPA in multivariable analyses included age (odds ratio [OR] 0.97), female sex (OR 1.86-2.01), moderate and vigorous leisure-time physical activity (OR 0.60-0.84), excessive drinking (OR 0.68-0.83), depressive symptoms (OR 1.28-1.96), anxiety symptoms (OR 1.63-2.45), sleep disturbance (OR 1.62-1.79), and arthritis/rheumatism (OR 1.32-2.18). Nonroutine nonmanual occupation (manual occupation as reference), body mass index, and smoking were independently associated with either any pain or PPA. CONCLUSION: This study provided preliminary information on the epidemiology of pain at baseline of the largest Latin American cohort on chronic noncommunicable diseases.