We evaluated reattached tooth fragments in terms of fracture resistance in endodontically treated maxillary premolars with one remaining wall. In total, 60 double-rooted, mature, human maxillary premolars with double canals were used. Ten intact teeth were served as a control group without any application. After endodontic treatment, the palatal wall of the teeth was cut parallel to the horizontal axis on the cementoenamel junction using a diamond disc. The teeth were randomly divided into five groups (n=10) and were restored as follows. Group 1: restored with composite resin; Group 2: palatal wall bonded to the teeth using adhesive resin, then restored with composite resin; Group 3: teeth restored as in Group 2, plus a post space prepared below 2mm from the palatinal cusp horizontally, then fibre post bonded using adhesive resin cement; Group 4: cavity enlarged to a depth of 2mm in the cusp of the palatal wall, then restoration performed as in Group 2; and Group 5: fibre post placed in the root canal and then restoration performed as in Group 1. Failure load testing was performed with a universal testing machine. Kruskal-Wallis and Conover's multiple comparison tests were used to analyse the data. In the restoration groups the highest median load at failure was observed in Group 4, in which cusp capping was used; the lowest was found in Group 3, in which a horizontal fibre post was applied with the reattached tooth fragments.