In this cross-cultural study of Puerto Rican and Texas physicians, we have tested the hypothesis that physicians' "humanistic attributes" are culturally related. Differences (P = .05 to .01) were found between mean responses of the two groups of residents to seven of the questions on the Totalitarian-Authoritarian-Dogmatism scale. Factor analysis determined that the two principal components of disagreement were medical political and ethical issues (P = .001) but not patients and their problems. Faculty responses provided additional evidence that customs and traditions influence how professional subgroups manifest attitudes: Puerto Rican faculty rated residents highest on "humanistic traits" if they self-reported high ethical scores, and lowest if they self-reported high radical scores; Baylor faculty reflected the "age-drag" on changing values by differing (P = .05 to .001) from both groups of residents on 38% of the questions. The difficulty of rating residents with a universal standard of "humanistic traits" is supported by the culture and age-specific differences reported herein.