Proxy agreement and internal structure of the Craig Handicap Assessment and Reporting Technique are documented, and the measure is compared with the Functional Independence Measure for a sample of stroke survivors. Thirty-eight former rehabilitation patients were assessed at follow-up (average time post-stroke was 6 mo) with the Craig Handicap Assessment, which was also answered separately under proxy instructions by caregivers (relatives or friends) who accompanied patients to the interview. Proxy instructions were for caregivers to answer as if they were the stroke survivor. Proxy agreement for the Craig Handicap total score was good (intraclass correlation was 0.77) and adequate for most of its scales. The physical independence, mobility, and occupation scales were highly intercorrelated, with low intercorrelations for social integration and economic self-sufficiency. Correlation between the handicap and Functional Independence Measure disability scores was around 0.50; when the handicap economic self-sufficiency scale was removed, this increased to 0.70. These findings are discussed within the context of the interrelationships among the educational level of the patients and both measures. With some modification, the Craig Handicap Assessment and Reporting Technique appears to be a useful tool for assessing outcomes in terms of the handicap status of stroke survivors.