Pulmonary embolism is one of the leading causes of mortality in pregnancy in the Western world. No clinical prediction models have been validated in pregnancy. As a result, any pregnant woman presenting with signs possibly consistent with pulmonary embolism is investigated radiologically. This study investigates whether using clinical prediction models for pulmonary embolism in pregnancy should be pursued in future prospective trials. The aim of this study was to retrospectively evaluate the Wells clinical prediction model and ventilation-perfusion scanning for pulmonary embolism in pregnancy. A retrospective study was performed on consecutive pregnant women who presented with suspected pulmonary emboli and underwent ventilation perfusion scanning at two tertiary institutions from 2007 until 2010. The clinical pretest probability was determined as likely or unlikely by two independent clinicians retrospectively using Wells-modified criteria. Scans were determined as normal, nondiagnostic or high probability for pulmonary emboli independently by two experienced radiologists. Disagreements were resolved by a third assessor independently. In 183 pregnant women, the pretest probability was determined as ‘pulmonary emboli likely’ in 76 (42%) and ‘pulmonary emboli unlikely’ in 107 (58%) of women. Scans were of high probability in four (2%), nondiagnostic in six (3%) and normal in 173 (95%) of women. This gives the pretest probability using Wells-modified criteria a sensitivity of 100% [95% confidence interval (CI) 0.4–1.0] and a negative predictive value of 100% (95% CI 0.96–1.0). A structured clinical model such as modified Wells criteria may be useful in pregnancy, but further prospective evaluation is required.