This report contains follow-up data on the effects of an anticoagulant-psychotherapy regimen in presenile and senile dementia. Over a two-year period, 49 such patients who were seriously ill were treated with a Coumadin-psychotherapy regimen; 34 (69 percent) improved (4 of them dramatically) and 15 (31 percent) did not improve or became worse. The underlying pathologic processes are discussed, with the rationale for therapy.