OBJECTIVES: To determine the amount of blood that should be discarded from a peripheral normal saline lock, a capped-off intravenous port, before a blood sample is obtained for determination of activated partial thromboplastin time from patients being treated with heparin. METHODS: A prospective, quasi-experimental design was used with 32 patients. A blood sample was obtained via venipuncture from each patient to serve as the control for that patient. The normal saline lock was flushed with 2 mL of normal saline. Four consecutive 3-mL blood samples were obtained directly from the normal saline lock, representing samples obtained after discard volumes of 0, 2, 4, and 6 times the dead space of the catheter and extension set (1.5 mL). Activated partial thromboplastin times for the venipuncture blood sample were compared with the times for the blood samples obtained from the normal saline lock. RESULTS: The only significant difference (P = .02) was that activated partial thromboplastin time was 15% higher in the blood sample obtained from the normal saline lock with no blood discarded than in the venipuncture blood sample. CONCLUSIONS: Nurses can obtain accurate measurements of activated partial thromboplastin time with blood samples obtained from normal saline locks by first discarding a volume of blood equal to 2 times the dead space of the catheter and extension set. Obtaining blood samples in this manner reduces patients' discomfort due to repeated venipuncture and diminishes blood loss.