Background: To identify multimorbidity patterns, by sex, according to sociodemographic and lifestyle in ELSA-Brasil. Methods: Cross-sectional study with 14,516 participants from ELSA-Brasil (2008–2010). Fuzzy c-means was used to identify multimorbidity patterns of 2+ chronic morbidities, where the consequent morbidity had to occur in at least 5% of all cases. Association rule (O/E≥1.5) was used to identify co-occurrence of morbidities, in each cluster, by sociodemographic and lifestyle factors. Results: The prevalence of multimorbidity was higher in women (73.7%) compared to men (65.3%). Among women, cluster 1 was characterized by hypertension/diabetes (13.2%); cluster 2 had no overrepresented morbidity; and cluster 3 all participants had kidney disease. Among men, cluster 1 was characterized by cirrhosis/hepatitis/obesity; cluster 2, most combinations included kidney disease/migraine (6.6%); cluster 3, no pattern reached association ratio; cluster 4 predominated co-occurrence of hypertension/rheumatic fever, and hypertension/dyslipidemia; cluster 5 predominated diabetes and obesity, and combinations with hypertension (8.8%); and cluster 6 presented combinations of diabetes/ hypertension/heart attack/angina/heart failure. Clusters were characterized by higher prevalence of adults, married and participants with university degrees. Conclusion: Hypertension/diabetes/obesity were highly co-occurred, in both sexes. Yet, for men, morbidities like cirrhosis/hepatitis were commonly clustered with obesity and diabetes; and kidney disease was commonly clustered with migraine and common mental disorders. The study advances in understanding multimorbidity patterns, benefiting simultaneously or gradually prevention of diseases and multidisciplinary care responses.