Social surveys have been the conventional means of evaluating the annoyance caused by transportation noise. Sampling and interviewing by telephone, mail, or in person are often costly and time consuming, however. Data collection by web-based survey methods are less costly and may be completed more quickly, and hence, could be conducted in countries with fewer resources. Such methods, however, raise issues about the generalizability and comparability of findings. These issues were investigated in a study of the annoyance of aircraft noise exposure around Brazil's Guarulhos Airport. The findings of 547 interviews obtained with the aid of Facebook advertisements and web-based forms were analysed with respect to estimated aircraft noise exposure levels at respondents' residences. The results were analysed to assess whether and how web-based surveys might yield generalizable noise dose-response relationships.