PURPOSE: To assess polypharmacy in a population with emphasis on regions. METHODS: We studied the individual-based data of all dispensed prescription drugs (DP) during a 3-month study period in Sweden 2006. As an indicator of polypharmacy, five or more (DP > or = 5) different drugs (substances) dispensed were applied. For analysis, we used comparisons of prevalence, correlation of prevalence of polypharmacy with different socioeconomic variables, and a novel weighted polypharmacy index. RESULTS: The national prevalence of polypharmacy, DP > or = 5, was 10.5% (inter-regional variation 9.1-12.1%). The regional variation in the prevalence of polypharmacy was largest for the age groups > or =90 (45.6-59.1%), 80-89 (46.1-53.4%) and 70-79 years (33.1-38.0%). The national prevalence of excessive polypharmacy, DP > or = 10, was 2.2% (inter-regional variation 1.9-2.6%). The regional variation in prevalence of excessive polypharmacy was largest for the age groups > or =90 (9.8-22.3%), 80-89 (11.4-17.1%) and 70-79 years (7.0-9.4%). We found a fairly strong positive correlation between polypharmacy and the age group > or =70 years (r = 0.84 for men and 0.71 for women). The novel weighted polypharmacy index indicated regional differences in the internal distribution of the prevalence of dispensed drugs for individuals with polypharmacy. CONCLUSION: Our findings indicate that the observed regional differences in the prevalence of polypharmacy partly can be explained by the regional age distribution in Sweden. The use of the novel weighted polypharmacy index indicated regional differences in drug therapy for individuals with polypharmacy.