Food products are a source of energy and essential substances but also of anthropogenic contaminants such as heavy metals. The aim of the study was to assess population health risks posed by contamination of food products with heavy metals, taking into account peculiarities of food preferences. An epidemiological study of actual nutrition of adult population of Samara region was conducted, the sample size was 1,856 people. At the first stage, using factor analysis, respondents' adherence to a certain model of food preferences was established; at the second stage, 5 homogeneous groups (clusters) of people with similar types of nutrition were formed using cluster analysis. The first cluster included individuals with maximum commitment to a high level of consumption of all studied foods; the second cluster was characterized by commitment to consumption of high-calorie foods such as baked goods, confectionery, sausages, potatoes, eggs, and cheese. Individuals from the cluster 3 showed a distinct preference for consumption of vegetables, fruit and dairy products. Individuals from the cluster 4 had no special preferences for any of the studied foods. The fifth cluster included people who had maximum preference for meat and meat products, smoked meats, pickles and salted fish. The content of cadmium, mercury, lead, and arsenic in food products was assessed via atomic absorption and photometric methods. The study relied on using social and hygienic monitoring data from the Samara Regional Rospotrebnadzor (Federal Service for Surveillance on Consumer Rights and Human Wellbeing) collection. Risk assessment of carcinogenic and non-carcinogenic effects was carried out in each of the five formed clusters taking into account modern methodological approaches. It was found that in all food clusters, the hazard coefficients for intake of contaminants in median concentrations and in the 90th percentile did not exceed permissible levels. In all clusters, the endocrine system was most at risk (HI = 1.68 ÷ 1.25). For all clusters, carcinogenic risk (for median concentrations) was created by arsenic both at the individual and the population level. The risk was the highest for people whose diets were characterized by high levels of consumption of high-calorie products. Cluster approach makes it possible to identify the most vulnerable groups of population in terms of risk burden for making managerial decisions and carrying out preventive measures.