One hundred unselected brain-damaged outpatients received a standardized battery of numerical tests (EC301) and, independently, a questionnaire on numerical activities in daily life (ADL). Comparisons between the two types of measurement were drawn from the scorings for different functional components of the calculation and number processing system. Results indicated high ranking correlations between the two instruments. The EC301 battery generally proved more powerful than the questionnaire in detecting the presence of mild-todiscrete impairments, but some aspects of numerical difficulties in nonaphasic patients were scored higher on the ADL questionnaire. (JINS, 1996, 2, 177–180.)