Morbid anatomical data have been examined in detail from 13 fatal diving accidents in which the primary cause of death was due to pulmonary damage. In addition, the observations on 6 cases of experimental barotrauma on cadavers are recorded. The information is supplemented by the incidental findings in the respiratory system of 19 professional divers killed by causes other than barotrauma, and 97 healthy, age-matched controls who died suddenly or unnaturally. In cases of pneumothorax resulting both from diving and experiment the constant feature was the presence of pleural adhesions and lung bullae, some of which were involved in causing tissue damage. A high proportion (up to 26%) of lungs not involved in barotrauma also showed pleural adhesions. In cases of barotrauma, the parenchymal damage is not consistently related to scars or fibrous tissue. The observations suggest that a high incidence of pleural lesions are not detected or detectable on routine medical examination which do not compromise the lung. However once a pneumothorax has occurred it is highly probable that a further incident may rupture another bulla and these appear to be multiple. The converse appears to be true with parenchymal barotrauma that is not consistently related to scar tissue. This would therefore suggest that in a proven and recovered case of barotrauma it should not necessarily debar further diving activity.