A synthetic pentadecapeptide, desacetylscotophobin analogue (DS 1-15), which differs from the natural product in terms of activity and in structure, has been tested in various behavioural tests. To determine the active core of DS 1–15, two other peptides corresponding to the residues 8–15 of the parent molecule (ZS 8–15 and HS 8–15) were tested in the same behavioural situations. These three peptides given in a single subcutaneous injection of 5 μg delayed extinction of a pole jumping avoidance response in an equipotent manner for approximately 7 days. In a passive avoidance situation, in which latency to enter a dark box was measured after animals had experienced mild shock 24 hr previously, avoidance latency was increased for several days following a single subcutaneous injection of 5 μg DS 1–15 and HS 8–15, while ZS 8-15 was much less active. None of these peptides affected passive avoidance latency in rats which had no previous shock experience in the dark box. Peptides were virtually inactive in a light-dark preference test and in an open field test. It is suggested that discrepancies in data with scotophobin and analogues as reported in the literature could well be the result of differences in aversive stimulation to which animals might have been unintentionally exposed.