Purpose to test the hypothesis that the effect of latent extinction in decreasing the number of trials to extinction would be dependent on the reduction or elimination of the secondary reward value of cues present in the goal box during the latent extinction period. Sixty male albino rats were used. All rats were trained in a T-maze with food reward in one of the goal boxes. After the learning series, the rats of Groups 1 and 2 were placed in the goal box for four 1-min. periods without the presence of food (latent extinction). In Group 1 the empty food cup was present in the goal box during latent extinction; in Group 2 the food cup was absent during latent extinction. Group 3 was not given latent extinction. Following the last latent extinction period, response extinction was begun. Each group was divided into two subgroups. In one subgroup the empty food cup was present in the previously correct goal box during the extinction trials, in the other the food cup was absent The extinction scores indicated that the secondary reward value of the food cup was reduced significantly more for Group 1 than for Group 2. No significant difference was found between the mean extinction score of Group 2 as compared with that of Group 3. Since, in addition, Groups 2 and 3 required significantly more trials to extinction than Group 1, the experimental hypothesis was supported.