Alexithymia is a hypothetical personality construct that has been associated with a variety of medical and psychiatric disorders. This article reviews a program of research evaluating the validity of the construct using a measurement-based, construct validation approach. For this purpose the Toronto Alexithymia Scale (TAS) was developed. In a series of studies the TAS demonstrated internal consistency, good test-retest reliability, and a stable factor structure theoretically congruent with the alexithymia construct. In separate tests of construct validity, the TAS correlated in a theoretically meaningful fashion with measures of other constructs. Criterion validity was supported by a study in which the TAS was able to discriminate between behavioural medicine outpatients designated as alexithymic and those designated as nonalexithymic on the basis of objectively rated structured interviews. In a normal adult sample, TAS scores were not related to sociodemographic variables or intelligence. These results provide considerable empirical support for the validity of the alexithymia construct. In addition, the TAS appears to be a psychometrically sound measure of alexithymia that may prove useful in testing the construct with psychiatric and medical patient populations.