Purpose: The purpose of this study was to investigate the marginal precision of computer numeric control- milled frameworks fabricated of grade 4 commercially pure titanium or cobalt-chrome alloy through digital technology and to compare them with conventional cast frameworks. Material and Methods: A titanium cast of a mandibular arch with six implant analogs was used as a master. The master cast was measured with a coordinate measuring machine. Fifteen rigid anatomic frameworks were created on the master cast in cast gold alloy and milled in titanium or cobalt-chrome material. The fifteen anatomic frameworks were measured in the same manner as the master cast. While the milled frameworks were measured once, at the end of the milling process, the cast anatomic frameworks were measured twice: immediately after the casting and divesting procedures and again after a technical adaptation procedure. Each anatomic framework was weighed. To compare the measurements obtained from each group of frameworks, descriptive statistics were calculated and one-way analysis of variance was performed, with values considered statistically significant at P .05. Results: The mean weight of the cast frameworks was 33.41 g, the cobalt-chrome frameworks weighed 18.12 g on average, and the titanium frameworks averaged 8.7 g. The mean values for three-dimensional deviation of the center point position for each group of frameworks were 261 µm (cast frameworks before adaptation), 49 µm (cast frameworks after adaptation), 26 µm (milled frameworks in cobalt-chrome), and 26 µm (milled frameworks in titanium). Conclusions: Within the limitations of this in vitro study, absolute passive fit cannot be achieved, regardless of material and fabrication technique. Anatomic milled frameworks fabricated in titanium or cobalt-chrome presented reduced center point deviation compared to cast frameworks. Titanium frameworks weighed less than cobalt-chrome and cast gold alloy frameworks.