Motlgolian gerbils received either continuous or partial (50%) reinforcement in acquisition of a shuttle avoidance response (Ra). In extinction, separate groups received CS termination at one of three delay intervals, 0, 5, or 10 sec, combined with shock omission. Removal of prompt CS offset produced rapid extinction of the Ra for both partial and continuous reinforcement animals. However, partially reinforced gerbils subjected to 5-sec delayed CS offset demonstrated a precipitous loss in Ra strength relative to their continuously reinforced . counterparts. These results were discussed in tenns of the interaction between the efficacy of the CS tennination contingency in extinction and acquisition reinforcement schedule.Although delay of CS termination following response completion has long been known to produce inferior avoidance acquisition relative t immediate CS termination (e. g., Kamin, 1956;Mowrer & Lamoreaux, 1942), the introduction of delayed CS termination has been employed as an extinction procedure only more recently. Thus, Owen (1963) demonstrated that delaying CS offset for 20 sec following an avoidance response (Ra) in extinction greatly decreased Ra strength compared to prompt CS termination. Katzev (1967) subsequently found that the ease of weakening the Ra varied directly as a function of delay duration (0-1 0 sec) following an Ra.Both Owen and Katzev omitted shock in extinction. The present author (Galvani, 1973a), on the other hand, assessed the effects of the CS termination contingency, following both partial and continuous reinforcement, in extinction in which shocks were unavoidable. These results showed that partial reinforcement (PR) in avoidance acquisition, in which nonreinforced trials involved the elimination of both CS termination and US avoid· ance, conSiderably enhanced the efficacy of the CS termination contingency in Ra maintenance. Specific· ally, a comparison of the relative strengths of association (w 2 statistic, see Hays, 1963, pp. 381-384) between delay of CS termination (0 vs. 5 sec) and extinction performance, following either PR or continuous rein· forcement (CR), showed that this relationship was much stronger in the case of PR-trained animals. Galvani (I 973a) speculated that the enhanced reinforcing capacity of CS termination following intermittent reinforcement might be attributable to its unique informational role, i. e., CS termination was the only reliable feedback signaling a successful Ra the animals experienced under the PR schedule.The research reported in the present paper was supported by Grant B04355-001 awarded to the author by the National Science Foundation.The present study was undertaken to further evaluate the effects of duration of CS termination delay on extinction performance following both continuous and intermittent reinforcement. Shock was omitted in extinction, since it is possible that the unavoidable·shock extinction procedures in Galvani (1973a) attenuated the effectiveness of CS termination. Assuming PRo avoidance training enhances the ...