The aim of the present study was to evaluate whether the anticipatory pupillary dilation response is a useful indicator for the examination of complex differential conditioning problems like patterning. A human fear conditioning procedure with six groups (n=20 each) was used to examine conditioned stimulus (CS) processing when a compound stimulus was reinforced, but not its elements (positive patterning) or when the elements were reinforced, but not the compound (negative patterning), as well as modifications in which the compound was replaced by either a new compound or by a new element. We found evidence for conditioning within 2 s after CS onset. Group differences in differential conditioning indicated systematic differences in CS processing due to different discrimination task difficulties. This study showed that the pupil response is a suitable indicator for human Pavlovian conditioning with high time resolution within and between trials.