This research investigated differences in task performance and perception under six non-time-varying ventilation-type background noise spectra with differing tonality. The results were related to five indoor noise criteria systems: noise criteria, balanced noise criteria, room criteria, room criteria mark II, and the A-weighted equivalent sound pressure level ͑L Aeq ͒. These criteria systems are commonly used in the U.S. building industry, but concerns exist over whether they are appropriate for all noise situations. Thirty test subjects completed three types of performance tasks ͑typing, reasoning, and math͒ and answered questions about their perception of the indoor environment under each noise condition. Results showed that performance scores did not change significantly across the six noise conditions, but there were differences in subjective perception. For example, perception trends for tonality, annoyance, and distraction changed based on the frequency and prominence of discrete tones in noise. However, these perceptual changes were not fully reflected in the criteria level or spectral quality ratings. Additionally, task performance was related to subjective perception but not to criteria level predictions. As a result, the authors suggest that the current criteria should be modified to account for the frequency and prominence of tones in background noise.