Objectives: Data have been collected on self-rated health using a single question on how individuals rate their health in cross-sectional surveys carried out in a large number of countries. Doubts have been expressed about the validity of this measure and this was the main reason to undertake the current study. Study Design: Data of 21 cross-sectional surveys were analyzed derived from the World Health Survey (WHS) carried out among adults in 2002-2003. Methods: We compared the single-item self-rated health measure with a multi-item health status index. Information on both types of measures was available from WHS. The multi-item index was constructed using data on functional limitations in daily activities. Results: The relationship of age with the multi-item health status index was linear while the relationship of age with self-rated health deviated from linearity in the younger and the oldest age groups. Both measures were compared with two criterion variables: life expectancy at age 20 and self-reported chronic conditions. The multi-item index was more strongly related to life expectancy and to chronic conditions than was the single-item self-rated health measure. Conclusions: The multi-item health status index could be a stronger predictor of mortality than the single-item self-rated health measure. It is recommended to rely in health surveys as much as possible on multi-item health status measures. Single-item self-rated health measures should continue to be used in situations where there are no other alternatives available, but researchers and policy makers should be aware of their limitations.