Response patterns derived from dichotomized (0/1) weekly CGI ratings conducted in antidepressant drug trials (Quitkin et al, 1984) were compared with those found in the pooled data from several randomized double-blind trials comparing the relative efficacy and tolerability low-dose flupenthixol im with that of three trieyclics (amitriptyline sr, imipramine, doxepine). Using the configurational frequency analysis (Krauth and Lienert, 1973), the postulated patterns could be rediscovered in our data apart from “early onset persistent patterns” which were less frequent in Quitkin et al's (1984) drug data. However, apart from this finding no “typical” patterns in terms of drug- or placebo-dependent response patterns could be detected in either the flupenthixol or Quitkin et al's (1984) data. It is concluded that there is little empirical evidence for the assumption of placebo- or drug related change- or response patterns. Moreover, theoretical aspects do not support the usefulness of such concepts.