Objectives To assess the fatigue behavior of a glass-ceramic and a resin composite milled with two different protocols in a 4-axis milling machine. Methods Lithium disilicate and resin composite crowns were milled using fine or extra fine mode in a 4-axis machine (n= 15). A full crown molar fiber-reinforced resin epoxy die was scanned and used as substrate. Respective surface treatments were applied, and the restorations were bonded with dual resin cement. Cyclic fatigue test was carried out (20 Hz, initial load 400 N, step size 50 N up to 1500 N, then increased to 100 N with 10,000 cycles per step) to assess the fatigue behavior up to the first crack. All specimens were submitted to a fracture load test until catastrophic fracture. Finite element analysis, topography, and fractography were also conducted. Bur wear was analyzed. Kaplan-Meier and two-way ANOVA tests were carried out to analyze fatigue behavior and fracture load, respectively. Results No significant difference in fatigue behavior was observed between the milling modes for both materials. Resin composite presented a higher survival rate until crack than lithium disilicate. No difference was found in fracture load among the groups. Extra fine milling mode promotes higher bur wear compared to the fine mode. Noticeable contact damage was observed in both materials. Significance Resin composite crowns exhibit superior fatigue behavior until crack detection compared to lithium disilicate, regardless of the milling mode. No significant difference in catastrophic fracture load was observed between the milling modes or the restorative materials.