Introduction: Histological subtype is an established prognostic factor in malignant pleural mesothelioma (MPM). We retrospectively investigated the accuracy of classifying histological subtype on diagnostic biopsies and examined the impact of different diagnostic procedures on the outcome. Methods: Consecutive patients with histologically confirmed MPM who underwent extrapleural pneumonectomy (EPP) from 1994 to 2009 were included. Patient records were reviewed, and the initial diagnoses of histological subtype were obtained. The archival EPP specimens were reviewed by a panel of pathologists. The histological subtype obtained at review was compared with the initial diagnosis. Results: Eighty-five patients underwent EPP. Two patients achieved a pathological complete response after neoadjuvant chemotherapy, leaving 83 patients to be included in this review. Different diagnostic methods were used before EPP: 81% thoracoscopy; 7% thoracotomy; 11% computed tomography-guided procedure; and 1% other. Patients determined to have an epithelial subtype (n = 64) at EPP were diagnosed correctly at initial diagnostic biopsy in 84% of cases, whereas patients considered to have a biphasic subtype (n = 19) at EPP were diagnosed correctly at diagnostic biopsy in 26% of cases. The sensitivity and specificity of diagnostic biopsy for epithelial MPM was 93% and 31%, respectively. The overall subtype misclassification rate was 20%. Biopsy by thoracotomy was most accurate in subtype classification (83%) compared with thoracoscopy (74%) and computed tomography-guided procedure (44%). Conclusions: The determination of histological subtype from a diagnostic biopsy is difficult due to sampling error, but an adequate specimen obtained from surgical biopsy increases the accuracy of subtype classification compared with radiological-guided biopsies.