We instituted a pharmacy-based surveillance and intervention program for heparin-induced thrombocytopenia (HIT). Surveillance occurred between August 1996 and July 1999. Platelet counts were monitored by pharmacists in patients receiving heparin 10,000 U/day or more. For patients with declining platelet counts, serology was ordered and clinicians were advised as appropriate. Outcomes of HIT in the surveillance group were compared with historical cases. During surveillance of 8,672 heparin courses, the incidence of HIT was 0.2%; however, the estimated rate in patients exposed to heparin for more than 4 days was 1.2%. Compared with historical HIT cases, the rate of thrombosis was reduced from 50% to 29% (p=0.39) during surveillance. The only patient factor associated with increased risk was the presence of cancer (p=0.03). This pharmacy surveillance method may have a role in improving outcomes of HIT.