Statement of problem The elastic property of Ti-Ni alloy may be a desirable property for cast removable partial dentures. However, little is known about the retentive properties of cast Ti-Ni alloy clasps. Purpose This in vitro study investigated the retentive force of various types of clasps during repeated cycles of placement and removal to determine whether Ti-Ni alloy clasps maintain their initial retentive force under varied conditions, including different retentive undercut depths and clasp size (thickness and width). Material and methods The test models were developed using a mandibular dentiform with a horizontal plane of occlusion. Two complete metal abutment crowns were made for the left second molar, differing only in retentive undercut depths. A total of 98 clasps (n=7) were fabricated, including 14 wrought wire clasps and 28 clasps from each of the following alloys: Co-Cr alloy (Biosil), Type IV gold alloy (Cast-4), and Ti-Ni alloy (TN-10). Clasps were made to engage one of 2 retentive undercut depths (0.25 mm or 0.75 mm) and were made in 2 sizes (0.8 mm or 1.4 mm). Each clasp was radiographically examined for casting defects and porosity. The force (N) required to remove the clasps was measured using a universal testing machine with a crosshead speed of 10 mm/min. After measuring the retentive force, a masticatory simulator was used to cycle the clasps on and off the metal crown 500 times, simulating the insertion and removal of a removable partial denture clasp. The cycling sequence was repeated 10 times, and retention force was measured after each sequence. The internal surface of the clasp arms was examined for evidence of metal fatigue using scanning electron microscopy. Data were subjected to 2-way and 4-way analysis of variance, followed by Scheffe's multiple comparison test (α=.05). Results Co-Cr alloy and gold alloy clasps in the 0.25-mm retentive undercut groups experienced a gradual decrease in retentive force measurements ( P<.001). In contrast, the Ti-Ni alloy clasps maintained a retentive force of approximately 1.8 N and 2.6 N for the 0.8 mm and 1.4 mm clasp groups, respectively. Although Ti-Ni alloy clasps' retentive force was reduced slightly after the first cycling sequence, it was negligible compared with other clasps tested. The wrought wire clasps also retained their retentive force until the final cycling sequence. A similar trend was found in the clasp groups engaging 0.75-mm retentive undercuts. Conclusion Although the end-point retention for all the clasps was similar, there was less change in the retentive force of the cast Ti-Ni alloy clasps after repeated cycling sequences of simulated placement and removal.