We sought optimal timing for heparin therapy in general surgery (GS) patients. From 2001 to 2008, 95 GS patients with documented thromboembolic events (TE) were identified and compared with matched controls (age, gender, type of operation, date of operation, malignancy, and body mass index [BMI]). Timing of heparin therapy, characteristics of TE or bleeding events, and risk factors for TE were collected. Mean age (57 years), BMI (33 kgM-2), gender (55% male), malignancy (53%), and duration of operation (204 vs 191 minutes, P = not significant) were similar in both groups. Peri-operative (within 24 hours) heparin administration (study 56% vs control 64%, P = 0.2) was no different. Preoperative therapy was more common in the control group (77% vs 51%, P = 0.001). The regression model showed a protective effect for heparin if given preoperatively (odds ratio = 0.37, P = 0.047) with no effect if started >10 hours from incision. Mean blood transfusion (97 and 106 mL) and hemorrhagic events (4.5% and 5%) were similar in both groups (P = not significant). Median (range) length of hospital stay and mortality was higher in TE cases [19 (0-201) vs 6 (0-66) days, 11 vs 2 mortality in 100-person-years (P < 0.05)]. Heparin administration before GS is associated with >2-fold reduction in TE. The optimal time to start heparin seems to be 1 to 10 hours before the time of incision.