BACKGROUND: When deep venous thrombosis is suspected, objective testing is required to confirm or refute the diagnosis. OBJECTIVE: To determine whether the combination of a low clinical suspicion and a normal D -dimer result rules out deep venous thrombosis. DESIGN: Prospective cohort study. SETTING: Three tertiary care hospitals in Canada. PATIENTS: 445 outpatients with a suspected first episode of deep venous thrombosis. INTERVENTIONS: Patients were categorized as having low, moderate, or high pretest probability of thrombosis and underwent whole-blood D -dimer testing. Patients with a low pretest probability and a negative result on the D -dimer test had no further diagnostic testing and received no anticoagulant therapy. Additional diagnostic testing was done in all other patients. MEASUREMENTS: Venous thromboembolic events during 3-month follow-up. RESULTS: 177 (40%) patients had both a low pretest probability and a negative D -dimer result. One of these patients had deep venous thrombosis during follow-up (negative predictive value, 99.4% [95% CI, 96.9% to 100%]). CONCLUSION: The combination of a low pretest probability of deep venous thrombosis and a negative result on a whole-blood D -dimer test rules out deep venous thrombosis in a large proportion of symptomatic outpatients.