Recent evidence indicates that the relationship between Type A behaviour pattern (TABP) and coronary heart disease (CHD) is dependent upon the method of measuring TABP. This suggests that the psychometric properties of TABP measures should be carefully investigated. This article examines one widely used TABP measure, the Bortner Scale, using data from 1320 working adults divided into three random samples. The reliability of the Bortner Scale as an overall TABP index is unacceptably low. However, further analyses indicate that, rather than reflecting a single dimension, the Bortner Scale contains two independent dimensions, one reflecting speed and the other reflecting competitiveness. The speed dimension was negatively related to job satisfaction and, to a lesser extent, positively related to anxiety and somatic symptoms, whereas the competitiveness dimension was positively related to job satisfaction. Implications for the use of the Bortner Scale are discussed.