Indiana University-Purdue University Indianapolis (IUPUI) / PURPOSE: To evaluate the fatigue failure load of distinct lithium disilicate restoration designs cemented on a chairside titanium-base (VariobaseTM for CEREC®, Straumann® LLC, USA) for restoring anterior implant restoration. MATERIALS AND METHODS: Left maxillary incisor restoration was virtually designed in 3 groups (n=10; CTD: lithium disilicate crowns cemented on custom-milled titanium abutments; VMLD: monolithic full-contour lithium disilicate crowns cemented on titanium-base; and VCLD: lithium disilicate crowns cemented on lithium disilicate customized anatomic structures then cemented on titanium-base). The titanium-base was air-abraded with aluminum oxide particles, 50 µm at 2 bars. Subsequently the titanium-base was steamed, air-dried and a thin coat of silane (Monobond Plus, Ivoclar Vivadent®, USA). All ceramic components were surface treated with hydrofluoric acid etching gel, follow by silanized, and bonded with resin cement (Multilink Automix, Ivoclar Vivadent®, USA). Specimens were fatigued at 20 Hz, starting with a load of 100 N (×5000 cycles), followed by stepwise loading up to 1400 N at a maximum of 30,000 cycles each. The failure loads, number of cycles, and fracture analysis were recorded. Data were statistically analyzed using one-way ANOVA followed by pair-wise comparisons (p < 0.05). Kaplan-Meier survival plots and Weibull survival analyses were reported. RESULT: For catastrophic fatigue failure load and total number of cycles for failure, VMLD (1260 N, 175231 cycles) was significantly higher than VCLD (1080 N, 139965 cycles) and CDT (1000 N, 133185 cycles). VMLD had higher Weibull modulus (11.6), demonstrating higher structural reliability. CONCLUSIONS: VMLD performed the best fatigue behavior when compared with the two other groups.