A one-page family history questionnaire was validated in two European
areas covered by population-based cancer registries. Information on ma
lignant tumor occurrence in first degree relatives was collected from
193 cancer patients in Trieste, Italy and from 64 in Basel, Switzerlan
d. They were then compared with the corresponding delta stored in the
registries' files. The sensitivity of the questionnaire was 85% (Tries
te) and 74%(Basel), the specificity was 97% in both studies and the ov
erall accuracy 95% (Trieste) and 94% (Basel). The questionnaire is rec
ommended for use in different geographical areas covered by population
based registries for comparative analyses of cancer related family hi
stories.