SYNOPSIS The traditional numerical scale used for measuring headache intensity was compared with an adjectival scale developed by Tursky (1976) designed to assess both the intensity and reactive components of the pain experience. As predicted Tursky's intensity adjectives correlated highly with the numerical ratings of intensity whereas all correlations involving the affective adjectives were significantly lower in magnitude. These results support the hypothesis that Tursky's procedure measures two different aspects of pain and suggest that research would profit from the incorporation of similar multi‐component self‐report measures.