The prerequisites for analyses of the medical, social and economic consequences of drug usage are in part available in Sweden. Hard data, though, are still fragmentary. Examples are given where various data sources and methods have been applied. It is suggested that feedback of drug utilization data should increase to create a more questioning attitude among prescribers. The concept of medical audit has to be better explained including the fact that individual-based registers are necessary tools in trying to assess the rationality of drug treatment. In the future such analyses should focus on everyday treatment of common disease entities such as hypertension, diabetes, dyspepsia and asthma. Long term medical and economical consequences of optimized pharmacological versus non-pharmacological treatment should be studied.