General measures of reaction to noise, which assess the respondent's perceived affectedness or dissatisfaction, appear to be more valid and internally consistent than more narrow measures, such as specific assessment of noise annoyance. However, the test-retest reliability of general and specific measures has yet to be compared. As a part of the large-scale Sydney Airport Health Study, 97 respondents participated in the same interview twice, several weeks apart. Test-retest reliabilities were found to be significant (p < 0.001) for two general questions and three specific "annoyance" questions. The general measures were significantly more valid for four of the six correlations (with activity disturbance), and more stable than the annoyance scales for five of the six possible test-retest comparisons. Amongst 1,015 respondents at Time 1, the questions regarding general reaction were more internally consistent than the questions regarding annoyance. Taken together, these data indicate that general measures of reaction to noise have superior psychometric properties (validity, internal consistency, and stability) compared with measures of specific reactions such as annoyance.