There is a paucity of theory pertaining to social surveys of annoyance and community noise. A new psychophysical model addresses data concerning resident annoyance in sites exposed to various maximum noise levels and different numbers of noise events. The best fits of this model point to the relevance of response variance for understanding the results obtained with category response scales. The argument identifies a number of psychological and physical variables that may influence annoyance reactions to community noise. Most importantly, the model suggests that the truncation inherent in category scales biases the responses, and that response variability may be critical in determining the detailed shape of the dose-response function. Regulatory decisions concerning annoyance and community noise should be more valid when based on a deeper understanding of the theoretical basis of the empirical data.