Results of epidemiological studies should be considered as part of the available evidence when evaluating drug treatment benefits in health technology assessment (HTA). Pharmaco-epidemiological databases can provide a broader understanding of the effectiveness of drugs in populations that are frequently underrepresented in clinical trials. Such databases are also useful to investigate drug safety with regard to socio-demographic and medical care-related indicators and hereby contribute to an optimal and targeted pharmacological therapy. Using examples from pharmaco-epidemiological asthma studies, the present article discusses associated difficulties in interpreting database results against the background of various sources of bias and proposes possibilities for integrating observational data into the HTA evaluation process. Researchers are challenged to engage in considerable efforts to develop a standardized inventory of epidemiological methods, e.g. for the pooled analysis of epidemiological data.