OBJECTIVES: The PROMIS system has undergone extensive validity testing in community populations, but has been much less studied in clinic patient populations.  We examined the patterns of correlation between the PROMIS Global Health Short Form measure and various scale scores on the SF-36.  METHODS: A series of patients (total N = 13) being seen for pre-surgical consultation were recruited from Henry Ford Hospital surgery outpatient clinic.  They were invited to complete a set of PROMIS CAT measures and the Global Health short form, and were also asked to complete the SF-36 survey.  SF-36 dimensions were scored on a 0-100 scale; Spearman correlation coefficients were calculated as the primary measure of agreement.  RESULTS: All of the correlations between PROMIS Global Health Score and SF-36 domain scores were in the expected directions, but none reached conventional levels of statistical significance except the correlation between PROMIS Global Health and SF-36 General Health (r = .55, p = 0.054).  Scores for essentially all the individual items within the PROMIS Global Health measure, though, were significantly correlated with both SF-36 General Health and SF-36 Bodily Pain.  Other individual items in the PROMIS measure correlated significantly with one or more SF-36 domain scores.  CONCLUSIONS: In this preliminary study of PROMIS measures in surgery patients, a pattern of general agreement between PROMIS Global Health and SF-36 domain scores was observed.  The modest levels of correlation, though, and the unexpected finding of higher correlations between the SF-36 scores and individual PROMIS scale items than between SF-36 scores and the total PROMIS scale score suggest the need for further exploration in a larger sample of clinic patients.