Two pigeons were trained on a six-key modified oddity-from-sample procedure. The stimuli were color pictures of birds, butterflies, and human faces. Initially, the third peck on the sample key (which presented one of three different bird pictures) lit only one comparison key. Every three additional pecks on the sample illuminated another comparison key. Fifteen sample pecks produced the maximum of five comparison stimuli. A peck on the comparison key that presented the nonmatching bird picture produced grain. Pecks on matching keys turned off all the comparison keys and repeated the trial. The birds learned to peck each sample until the nonmatching comparison stimulus was produced, and then to peck that key. After acquisition (70%-90% accuracy), the three bird stimuli were replaced by a new set of three bird pictures. Subsequent phases provided new sets of bird, butterfly, and human face stimuli. Both birds showed transfer of oddity learning to the novel samples. The data suggest that the birds may have been engaging in conceptualtype oddity learning, rather than learning discrete five-key discriminations or a series of two component chains.