Polypharmacy is commonly defined based on the number of medications taken concurrently using standard cut-offs, but several studies have highlighted the need for a multidimensional assessment. We developed a multidimensional measure of polypharmacy and compared with standard cut-offs. Data were extracted for 2141 respondents of the 2007 Prescription Drug Survey, a sub-study of the Health Retirement Study. Latent classes were identified based on multiple indicators of polypharmacy, including quantity, temporality and risk profile. A four-class model was selected based on fit statistics and clinical interpretability: ‘High risk, long-term’ (Class 1), ‘Low risk, long-term’ (Class 2), ‘High risk, short-term’ (Class 3), and ‘High risk for drug interactions, medium-term, regular’ (Class 4). Classes differed regarding sex, cohabitation, disability and multimorbidity. Participants in the ‘low risk’ class tended to be male, cohabitating, and reported fewer health conditions, compared to ‘high risk’ classes. Polypharmacy classes were compared to standard cut-offs (5+ or 9+ medications) in terms of overlap and mortality risk. The three ‘high risk’ classes overlapped with the groups concurrently taking 5+ and 9+ medications per month. However, the multidimensional measure further differentiated individuals in terms of risk profile and temporality of medication taking, thus offering a richer assessment of polypharmacy.