There are several methodical problems that have to be coped with when testing psychotropic drugs in psychiatric clinics. In this paper some relevant problems have been discussed which are encountered during that part of a clinical trial, starting with the assessment of the initial psychopathological state of the patients and ending with the statistical analysis. Especially the AMP-System was considered. Particularly the following points have been dealt with: comparison of different methods for the assessment of psychotropic effects, special problems of rating scales, studies on the AMP-system concerning psychopathological and methodical aspects, conversion and processing of AMP-data, general considerations on the statistical analysis of drug trials.