The use of 3D technologies is progressing in the dental field. However, little is known about the biomechanical behavior of the additive manufacturing of full-arch fixed dental prostheses (FAFDPs) for the establishment of clinical protocols. We investigated the influence of three CAD/CAM technologies: milling (control), Selective Laser Melting (SLM) and Electron Beam Melting (EBM) for FAFDP manufacturing. Also, the effects of ceramic veneer and spark erosion on marginal misfits of FAFDPs, the stability of prosthetic screws, strain and stress on the implant-supported system, as well as the effect of chewing simulation on screw stability were evaluated. Fifteen Ti–6Al–4V alloy FAFDPs were obtained by means of CAD/CAM systems: milling, SLM and EBM (n = 5/group). The marginal misfit was analyzed according to the single-screw test protocol. Screw stability was analyzed by screw-loosening torque. Strain-gauge analysis investigated the strain on the mini-abutment analog, and photoelastic analysis investigated the stress on the peri-implant region. Subsequently, all frameworks underwent ceramic veneer and spark erosion procedures. Marginal misfit, screw-loosening and strain and stress analyses were assessed after each evaluation time: initial, ceramic veneer and spark erosion. Finally, all prostheses were subjected to 106 mechanical cycles (2 Hz/150 N), and screw-loosening was re-evaluated. Data were subjected to two-way ANOVA for repeated measures, and the Bonferroni test as a post hoc technique (α = 0.05). At the initial time, the milling group presented the lowest marginal misfit (p < 0.001). Ceramic veneer did not alter marginal misfit for all groups (p > 0.05); spark erosion decreased the misfit values for the SLM and EBM groups (p < 0.05). Evaluation time did not alter screw-loosening values for all groups (p = 0.191), although the milling group presented the highest screw-loosening values (p < 0.05). Ceramic veneer and spark erosion reduced strain in the components regardless of the manufacturing technology used (p < 0.05). The milling group presented the lowest stress values regardless of evaluation time (p = 0.001), and lower stress values were found after spark erosion regardless of the manufacturing group (p = 0.016). In conclusion, although milled frameworks exhibited the best biomechanical behavior, frameworks manufactured by additive technologies presented acceptable values of screw-loosening torque, strain and stress. Ceramic veneer did not negatively interfere in the biomechanical tests of the study, and clinically acceptable marginal misfit was achieved after spark erosion. Therefore, such 3D printing technologies seem to be feasible for the manufacturing of full-arch implant-supported frameworks.