The purpose of this study was to test the actual resistance against fracture of a crown fragment fractured and reattached using three different techniques of preparation. Forty bovine incisors were randomly assigned into four groups of which the first was the control group. The elements of the three experimental groups were all cut at a fixed distance of 3 mm for the incisal margin on the buccal surface of the crown in a plane normal to the buccal surface itself, and subsequently re-bonded using an adhesive system. After the reattachment, on the teeth of group 2, a circumferential chamfer was performed along the fracture line; on the teeth of group 3 a groove called ‘overcontour’ was made also along the fracture line; the teeth in group 4 received a chamfer on the buccal surface and an overcontour on the lingual surface. Finally, all the elements were encompassed in chalk blocks and the models were mounted on the 858 Mini Bionix to perform a fatigue load test. A force was applied on the buccal surface of each tooth at 1.5 mm from the incisor margin, with a velocity of 1 mm min−1, through the use of a steel prick. All the elements of groups 2, 3 and 4 demonstrated a lower resistance to fracture in comparison with the elements of the control group. The teeth of group 2 showed a resistance to fracture equal to 36.1% of the resistance of the elements in the control group; in group 3 the resistance was equal to 50.2%, while in group 4 the resistance reached 55.9%. The difference in resistance between group 3 and group 4 did not result statistically significant (P = 0.82). Reattachment of coronal fragments does not restore the resistance to fracture to an equal level of the intact teeth, in a static test. The different techniques of preparation significantly modify the resistance to the fracture of a re-bonded fragment. The technique of the circumferential chamfer produced results that were less favourable in terms of resistance against fracture.