This brief report summarizes a study to identify primary bound conditions of sound level selection as a first step in collecting base-line data for evaluating selective listening performance in infants with known or suspected hearing loss. Ten normal 9 to 22 month old infants in their home cribs played with an automated operant "toy" that allowed them to choose between programs of nursery songs at two different loudness levels. In a two-phase record with more than 60,000 seconds of listening response time, the infants showed a highly significant preference for the louder feedback when the loudness levels differed by about 30 decibels. No preference pattern was shown when the difference in loudness levels was only 10 decibels. (Author/JS)