The use of signal detection with children provides a precise method for investigating auditory system parameters. This study investigated the use of signal detection with normal hearing children ages 6–12. A yes‐no procedure was designed such that a child could easily learn the task. A trail was initiated by the child pressing a single manipulandum. After initiation of the trial a 1‐sec delay occurred. A “listen” light then occurred simultaneously with a randomly presented (0.50 probability) 1000‐Hz pure tone. Tonal duration was three seconds. The child's task was to release the manipulandum if the tone was perceived. If the tone was not perceived the child was instructed to continue pressing the manipulandum. Tangible reinforcement was automatically dispensed for hits. Using this procedure a dense distribution of latencies were recorded for hits and false alarms. By redefining “yes” responses at 50‐msec intervals different hit and false‐alarm probabilities were obtained. These probabilities were used to construct ROC curves. This method generated a large number of points on the ROC curve with great economy of experimental time. Several stimulus intensities were used to generate functions relating intensity withd′.