The purpose of this paper is to discuss the results of a direct comparison of two instruments, the SF-12 health survey and the EuroQol Group's EQ-5D, in a sample drawn from the adult US population. The data were collected via a survey instrument mailed to 1,800 subjects in the USA. The instrument contained the EQ-5D valuation questionnaire and the items of the SF-12. In addition, the subjects were asked to provide demographic information, complete a depression screening item and to indicate if they had any of seven common chronic conditions. The usable response rate was 23.7%. Patients indicating a health problem on the EQ-5D had significantly lower mean SF-12 component scores (i.e. MCS-12 and PCS-12) for all dimensions. As was expected, the relationships were stronger between the EQ-5D functional dimensions and the PCS-12 and between the MCS-12 and the EQ-5D anxiety/depression dimension. The EQ-5D visual analogue scale (VAS) scores were positively correlated with both component scores; r = 0.55 for PCS-12 and r = 0.41 for MCS-12. The results of this investigation provide support for the validity of both the EQ-5D and the SF-12 as descriptive measures of health status. The measurements by both instruments behaved in patterns that were consistent with recognized sociodemographic differences in health status. The limitations of the dimensional structure of the EQ-5D were somewhat overcome by the use of the VAS, which may also be possible with a constructed index score based on explicit values derived from a general population. The SF-12 appeared to be more sensitive to differences associated with less severe morbidity.