This study evaluated whether variations in the structure of minimal versus maximal opposition treatments would result in empirical differences in phonological learning. Subjects were 3 children who excluded at least six sounds from their pretreatment phonetic and phonemic inventories. An alternating treatments design in combination with a staggered multiple baseline across subjects was used to evaluate differences in learning the two types of oppositions. Results indicated that treatment of maximal oppositions led to greater improvement in the children's production of treated sounds, more additions of untreated sounds to the posttreatment inventory, and fewer changes in known sounds than treatment of minimal oppositions. Moreover, individual differences suggested that phonological learning was enhanced not only by the number but also by the type of distinctions being taught. A potential sequence of relative clinical effectiveness was proposed such that teaching multiple and major class distinctions greater than teaching multiple distinctions greater than teaching few distinctions among sounds.