A survey of noise levels and noise annoyance was conducted along 19 urban and suburban roadways in Brisbane, Sydney and Melbourne. An average of 43 residents were surveyed along each roadway with a total sample of 818. Traffic volumes on the roadways ranged from 4000 to 57000 vehicle/ day. It was found that a succinct measure of the noise effects was provided by the residents self rated score on a simple seven point, annoyance scale. However no presently available noise level scale was found to correlate substantially with individual respondents' noise annoyance scores. It was confirmed that people with different noise annoyance susceptibilities had different responses to noise, but the hypothesis that different susceptibility groups had differently graded responses to increasing noise exposure was not supported. The best predictor of group annoyance scares was found to be the number of 'heavy vehicles', and this was felt to be a proxy measure of the number of 'noisy vehicles' using the roadway. (A)