Multilevel studies have gained importance for highlighting social inequalities in health. These associations have been reported previously in diseases such as arthritis and chronic pain. We conducted a cross-sectional study using multilevel analysis to identify individual and contextual factors associated with the variation of prevalence of osteoarthritis (OA) in the Mexican population. The sample included 17,566 individuals of which 10,666 (60.7 %) were women. The relationship between individual and contextual factors and OA were analyzed with a multilevel strategy. From the total population, 1,681 individuals had OA. Multilevel analysis showed that individual variables such as female gender (odds ratio (OR) = 1.3, 95 % confidence interval (CI) 1.1, 1.4), age range 55–65 years (OR = 1.6, 95 % CI 1.3, 2.0), musculoskeletal pain in the last 7 days (OR = 2.6, 95 % CI 2.3, 3.0), and use of pain treatments (OR = 1.4, 95 % CI 1.2, 1.7) were associated with OA. At the regional level, the Social Gap Index (SGIx) was associated with the diagnosis of OA (coefficient 0.5, 95 % CI 0.2–1.1). The SGIx contextual variable was positively associated with the regional prevalence of OA and the variation in prevalence of OA in different regions. The larger the social gap, the greater the variation in OA prevalence. These factors were independently associated with the prevalence of OA: female gender, pain intensity, physical limitation, and the use of pain treatments were individual variables associated with OA. The association between OA prevalence and regional variations with SGIx reflects inequities in health provisions that should be considered in health programs.