This study examined changes in infants' performance in detecting tones as a funtion of time. Stimuli were presented in blocks of fixed frequency and intensity, rather than in random order. The findings of earlier studies that an observer can reliably tell from a 1-month-old's behavior when a pure tone is being presented, and that reinforcement of infant responses to tones improves the observer's performance, were replicated. Results showed that observers performed as well in the fixed stimulus condition as in the random order condition. However, there were frequency-specific changes in performance over the course of a test session. References to three articles and numerous graphs are included. (BC)