In an observational study using heparinase-modified thrombelastography, we investigated the percentage of elective cardiothoracic surgical patients receiving low-dose unfractionated heparin (5000 IU 12 hourly subcutaneously) who had a demonstrable systemic heparin effect. Blood samples were obtained at induction from 40 adult elective cardiothoracic surgical patients who had received 5000 IU unfractionated heparin subcutaneously within six hours. Simultaneous kaolin and heparinase-modified thrombelastographies were run on all samples. Fourteen patients (35%; 95% CI: 20 to 50%) had a demonstrable heparin effect (defined as a kaolin thrombelastography R time >25% longer than the heparinase-modified control). Their mean +/- SD kaolin thrombelastography R time was 13.6 +/- 5.9 minutes (normal range 4 to 8 minutes) vs. 7.1 +/- 2.0 minutes for the heparinase-modified controls. In 10 patients the thrombelastography R times were >50% longer and in four patients >100% longer than their respective heparinase-modified controls. In a post hoc analysis, there was little correlation between the extent of the prolongation and patient age (r = 0.02), weight (r = -0.31), preoperative creatinine (r = -0.17), or time since administration of heparin (r = 0.14). These results indicate that about one third of patients who have received low-dose unfractionated heparin subcutaneously within six hours have a demonstrable heparin effect. The potential for this effect should be considered if central neural blockade is planned.