OBJECTIVES: Databases provide aggregated and confounded data regarding management and cost of pulmonary embolism (PE). The objective of this study was to investigate, via patient chart analysis, the average and the variation in medical resource use strictly related to PE patients, and to find explanatory factors for the possible variation. METHODS: Patients with confirmed PE (n = 54) were randomly selected from 5 Belgian centres. The centres were representative for size and region. The charts were analysed in detail, based on a standardised case record form. Only medical resource use related to the treatment of presumed and confirmed PE was included in the analysis, with a time horizon from the day of suspicion up to discharge. RESULTS: On average, the management of PE is rather consistent across centres. The total average cost of PE was equal to 3,394 Euro (95% CI = 2,762 – 4,027 Euro). Still, large inter-individual differences are found in the intensity of diagnosis and follow-up. More in particular, a large variation in hospital duration is found, (average = 14.6 days, Std. Dev. = 9.3), but this effect was not explained by the type of centre nor the region. The only two factors that are significantly related with hospital duration in a multivariate regression analysis are the number of days of heparin use and the intensity of diagnostic follow-up, as expressed by the number of VQ-scans. CONCLUSIONS: The management of PE, from suspicion until discharge, is rather consistent, however, large inter-individual variation in hospital stay is observed. Interestingly, the length of hospitalisation is related to the duration of heparin use. This perhaps opens perspectives for anticoagulation that can be administered ambulatory. A patient chart review has the advantage over database research that only resource use strictly related to the investigated condition can be identified.