The International Classification of Impairments, Disabilities and Handicaps (ICIDH) model proposes that there are three consequences of disease, impairment (I), disability (D) and handicap (H) and that they are sequentially related. This paper examines first, whether I, D and H can be measured independently and second, whether there is support for the sequential or causal relationship between the three constructs. Cross-sectional data from a representative sample of 101 disabled adults and longitudinal data from 108 myocardial infarction (MI) and 68 Stroke patients were used. Standard measures of I were used for each clinical condition. Measures of D and H were derived from the British version of the Sickness Impact Profile (SIP) and additional measures of D were available for the Stroke group. Judges classified SIP items according to ICIDH definitions of D and H. Correlation, Confirmatory Factor Analyses and Path Analyses were used to examine the main hypotheses. Valid measures of D and H were derived and there was evidence of separation of the three constructs for the Stroke patients but not for the other groups. For both Stroke and MI, I was not predictive of D and H. For Stroke, the best path model included a path from D to H, but not from H to D and this was found for self-report and performance measures of D. Using these measures, the ICIDH model was supported in that D predicted H for stroke, but there was no support for a path between I and D or between I and H. Further it was not always possible to distinguish the three constructs. Possible limitations in the measures and in the ICIDH model as a testable scientific model are discussed.