OBJECTIVES: To determine the population risk stratification based on the cost of care (health resource use) in patients with chronic diseases in primary health care (PC). METHODS: Multi-center observational design. We included all patients from 6 centers of PC that demanded assistance in 2010 managed by Badalona Serveis Assistencials SA (health organization). The risk population was defined beginning from the complexity (co-morbid chronic [CC]) and fragility (socio-demographic and clinical criteria). Main measures: services (medical, paediatric), chronic co-morbidity (CC) and direct cost model. From a group of experts identified the different chronic conditions and population risk levels: Level 1 (no CC), level 2 (1-2 CC), level 3 (3-4 CC) and level 4 (≥ 5 CC). Fixed (operation) and variable costs were considered. Statistical analysis: linear regression model (coefficient of determination [R2], dependent variable: health care costs) and principal components, p<0.05. RESULTS: We included 83,090 patients, mean age 40.9 years, women: 53%. The total cost was 56.1 million / EUR. The average / unit cost: 675.3 euros. The cost for drugs was 41%. Stratification levels: level 1 (N = 36,859, 44.4%, €283.9), level 2 (N = 32,644, 39.3%, €694.8), Level 3 (N = 10019, 12.1%, €1461.6, and level 4 (N = 3568, 4.3%, €2331.2). Musculoskeletal diseases (38.1%), mental (31.6%) and cardiovascular (30.4%) were the most frequent, p <0.001. Predictive model (R2): age = 23.4%, age-sex = 24.1%, age-sex-CC = 41.8% (medical: 47.9%; Paediatrics: 15.1%, p <0.001). It details the complexity and fragility of the patients for each level of stratification and clinical services. CONCLUSIONS: The CC is associated with increased healthcare costs. The number of co-morbidities explains much of the costs. Knowledge of the risk / complexity / fragility of the patients should allow preventive intervention strategies.