Results are reported for the effects of sodium pentobarbital on (a) zero delay matching-to-sample, (b) simultaneous oddity, and (c) variable delay matching-to-sample. In all these situations there are decrements in accuracy with increasing dose levels. The oddity performance proved to be considerably less sensitive to the drug effects than zero-delay matching. In variable delay matching there was no clear evidence for any relationship between the magnitude of the drug effect and the length of the delay interval.