Eighty-eight rats, 40 control and 48 experimental, were trained on a new type of T-maze apparatus in which it was possible to control intra- and extra-maze cues. Half of the control and half of the experimental were given 50 percent reinforcement and received one food reinforcement daily for 12 days. The other half of each were given 100 percent reinforcement and received 2 food reinforcements daily for 6 days. In the control groups no attempt was made to control the effects of secondary reinforcement, whereas in the experimental groups the conditions were arranged so as to minimize such effects. No significant differences were found between partial and continuous reinforcement in the control groups. But in the experimental groups the learning of the rats receiving 100 percent reinforcement was significantly superior to that of those receiving 50 percent reinforcement. The results are interpreted in terms of the operation of secondary reinforcement. (PsycINFO Database Record (c) 2006 APA, all rights reserved)