A trend towards increased utilisation of polypharmacy in clinical psychiatry has been observed recently. Among the factors contributing to this are the emerging predominance of drug treatment for psychiatric disorders, newer therapeutic agents, attempts to alleviate symptoms in patients who only partially respond to treatment, and the over-interpretation of research data.Problems with polypharmacy include the potential for drug interactions, the unknown safety of newer medicines, and the use of superfluous medication mixtures when one drug might suffice. The complexities inherent in the existing research methodology limit the study of polypharmacy. We suggest a new paradigm for researching multidrug regimens, modelled on that of health services research.