This study analyzed the degree of passive and vertical fit achieved in frameworks using either cobalt-chromium (Co-Cr) or titanium (Ti) implant-supported fixed partial dentures ( FPDs) fabricated with a CAD/CAM milling technique. 33 3-unit FDPs, 17 of Co-Cr metal alloy (test group) and 16 of Ti (control group), were manufactured with two implants by copy milled technology. Optical microscopy was used to measure passive fit (PF) and vertical fit (VF) in all frameworks. The PF was evaluated by using the Single Screw test and the VF with the screws tightened at 20 Ncm. Descriptive and inferential analysis were performed to evaluate statistically significant differences in the tested groups for each fit. Brunner-Langer models were applied to assess potential material and implant area effects on the measurements. An ANOVA test was performed to estimate both main effects and interactions. The average PF values in the screwed implant were 4.43 ± 0.52 µm for Ti and 5.50 ± 1.61 µm for Co-Cr and in the non-screwed implant 5.59 ± 1.32 µm in the group Ti and 6.25 ± 1.55 µm the Co-Cr group. In this last implant, it was not observed statistically significant differences between both types of alloy (p = 0.178) nor between zones. Ti control group exhibited a significantly better VF than Co-Cr (p = 0.046) in the screwed implant but there were no differences in the implant not screwed. The VF in the non-screwed implant was better in lingual than in buccal zone. The PF and VF measurements observed in Co-Cr frameworks are clinically acceptable. 3-unit implant supported FPDs made with Co-Cr alloy using milling technique showed similar dimensional accuracy than those obtained with Ti.