The subjective concept of health included in the self-rated health process allows identifying factors associated with good self-rated health in populations. This cross-sectional population-based study in 2007-2008 focused on self-rated health and associated factors among adults (n = 1,516) in Rio Branco, Acre State, Brazil. The multivariate analysis used Poisson regression, with fair/poor self-rated health as the dependent variable. Differences in gender, age group, level of schooling, and income negatively affected individuals' self-rated health. After adjusting for age, gender, and self-reported illness, the estimated risks for fair/poor health were higher for the unemployed, obese, and those with some self-reported illness. The results suggest that differences in living conditions for adults in Rio Branco are associated with precarious self-rated health.