Adults with Prader-Willi syndrome learned the conditional relations necessary for the formation of two equivalence classes under four conditions: (a) nondifferential, nonedible outcomes; (b) nondifferential, edible outcomes; (c) differential, nonedible outcomes; and (d) differential, edible outcomes. Tests for transitive relations revealed superior performance when the two differential outcomes procedures, in which a distinct reinforcer was associated with each stimulus set, were used during teaching. Performance on test trials following nondifferential outcomes training was better when edible outcomes were used during teaching for 4 of the 5 participants. An enhancement of performance on the derived relations separated by two or three nodal stimuli was seen when a differential outcomes procedure was used to teach the baseline conditional relations.