Cynomolgus monkeys (Macaca fascicularis) were assessed for their ability to associate visual stimuli with food reward. They learned a series of new 2-choice visual discriminations between colored patterns displayed on a monitor screen. The feedback for correct choice was the delivery of food. In order to promote associative learning between the visual stimuli and the incentive value of the food reward, reward delivery was not accompanied by any distinctive visual feedback on the display screen. The rate of learning new problems was assessed before and after surgery in a total of 16 monkeys. Three groups of 3 monkeys received bilaterally symmetrical ablations in either the amygdala, the mediodorsal nucleus of the thalamus, or the ventromedial prefrontal cortex. All these groups showed a severe postoperative learning impairment. Seven other animals were given a unilateral ablation in 1 of those 3 structures and a second unilateral ablation, in each case contralateral to and different from the first, in order to produce 2 groups in which a putative amygdalo-thalamo-prefrontal pathway had been disconnected by crossed unilateral lesions. One disconnection group had lesions in the amygdala and ventromedial prefrontal cortex; the other had lesions in the amygdala and the mediodorsal nucleus of the thalamus. The disconnection groups showed a significant impairment, but the effect of the disconnection surgeries was significantly milder than the effect of any of the 3 bilaterally symmetrical lesions. Therefore, symmetrical bilateral lesions in either the amygdala, the mediodorsal nucleus, or the ventromedial prefrontal cortex produce similar impairments in the present task, implying that these structures are functionally related to each other; but the relatively mild effect of disconnecting these structures from each other argues against the hypothesis that they are serial stages in a single, tightly linked functional pathway.