Longitudinal studies on tooth mortality are rare but the reasons for tooth mortality have been studied by several authors. The aims of this investigation were to study the reasons for and incidence of tooth mortality in an earlier described Swedish population of 200 patients on two occasions with an interval of 5-7 years, and to see if the reasons for extractions were correlated to posts, crowns or endodontic status, respectively. 197 (4.0%) of the 4889 teeth registered at the first examination were lost during the interval. 65 (33%) of the lost teeth were endodontically treated. 44 (68%) of these were registered as having one or more root with a root-filling ending more than 2 mm from the apex and 29 (45%) were judged to have an improper seal. In 93 of the 197 lost teeth it was possible to find the reason for extraction from the patient chart. Based on distribution of selected variables, these 93 teeth seemed to be representative of the whole group of 197. It was concluded that tooth losses were evenly distributed in the different age-groups and that above all molars but also premolars were lost more often than teeth in the frontal region. Furthermore, endodontically treated teeth were lost more often than other teeth and the quality of the root-filling affected the risk for losses while crowned teeth did not run a higher risk of being lost than teeth without crowns. Finally, it was found that caries, including pulpitis and apical periodontitis, was the main reason for tooth extractions.