Summary. Background: Multiple-detectors computed tomographic pulmonary angiography (CTPA) has a higher sensitivity for pulmonary embolism (PE) within the subsegmental pulmonary arteries as compared with single-detector CTPA. Multiple-detectors CTPA might increase the rate of subsegmental PE diagnosis. The clinical significance of subsegmental PE is unknown. We sought to summarize the proportion of subsegmental PE diagnosed with single- and multiple-detectors CTPA and assess the safety of diagnostic strategies based on single- or multiple-detectors CTPA to exclude PE. Patients and methods: A systematic literature search strategy was conducted using MEDLINE, EMBASE and the Cochrane Register of Controlled Trials. We selected 22 articles (20 prospective cohort studies and two randomized controlled trials) that included patients with suspected PE who underwent a CTPA and reported the rate of subsegmental PE. Two reviewers independently extracted data onto standardized forms. Results: The rate of subsegmental PE diagnosis was 4.7% [95% confidence interval (CI): 2.5–7.6] and 9.4 (95% CI: 5.5–14.2) in patients that underwent a single- and multiple-detectors CTPA, respectively. The 3-month thromboembolic risks in patients with suspected PE and who were left untreated based on a diagnostic algorithm including a negative CTPA was 0.9% (95% CI: 0.4–1.4) and 1.1% (95% CI: 0.7–1.4) for single- and multiple-detectors CTPA, respectively. Conclusion: Multiple-detectors CTPA seems to increase the proportion of patients diagnosed with subsegmental PE without lowering the 3-month risk of thromboembolism suggesting that subsegmental PE may not be clinically relevant.