The influence of casting and masticatory simulation on marginal misfit and strain in multiple implant-supported prostheses was evaluated. Three-unit screw retained fixed dental prosthesis (FDP) and screw retained full-arch fixed dental prosthesis (FAFDP) frameworks were made using calcinable or overcasted cylinders on conical dental implant abutment. Four groups were obtained according to the cylinder and prosthesis type (n=10). Frameworks were casted in CoCr alloy and subjected to strain gauge analyses and marginal misfit measurements before and after 10(6) mechanical cycles (2 Hz/280 N). Results were submitted to ANOVA, Tukey's HSD and Pearson correlation test (α=0.05). No difference was found on misfit among all groups and times (p>0.05). Overcasted frameworks showed higher strain than the calcinable ones (FDP - Initial p=0.0047; Final p=0.0004; FAFDP - Initial p=0.0476; Final p=0.0115). The masticatory simulation did not influence strain (p>0.05). No correlation was observed between strain and misfit (r=0.24; p>0.05). In conclusion, the marginal misfit value in the overcasted full-arch frameworks was higher than clinical acceptable data. It proved that overcasted method is not an ideal method for full-arch prosthesis. Overcasted frameworks generate higher strain upon the system. The masticatory simulation had no influence on misfit and strain of multiple prostheses.