Physiological and behavioral studies in animal models have revealed that information from the different senses can be used synergistically to enhance the detection of objects and events. Although a great deal of evidence exists which demonstrates the synergistic use of multisensory cues in human behavior and perception, there is conflicting evidence about whether such interactions can be used to aid in stimulus detection. To examine this issue, we had volunteers indicate the presence or absence of a brief, low-intensity sound that was either presented alone or paired with a simultaneous light in a one-interval signal detection task. In the first experiment, the task-irrelevant light was found to enhance the detectability of the sound, but also to increase the subject's willingness to report the presence of a sound (i.e., response bias). In the second experiment, designed to eliminate this response bias, we found an improvement in stimulus detectability in the absence of any change in bias. These findings demonstrate the presence of a significant multisensory-mediated gain in stimulus detection in human subjects.