To compare the dimensional accuracy of implant framework castings from an argon vacuum casting machine with those from a centrifugal casting machine.Three 4 x 10-mm external hex-type implants (3i/Implant Innovations) were embedded in an acrylic resin block 7 mm apart, with a 2 mm offset of the middle implant. Eight reference points were marked on the implant collars. Twenty implant bar frameworks were waxed with UCLA abutments, invested with a ringless system, and subjected to the same thermal cycle. Ten wax patterns were cast in gold alloy using an oxygen-propane torch and centrifugal casting system; 10 were cast using an argon vacuum casting machine (KDF; Denken). The White 1-screw technique was applied after sequentially tightening the mesial and distal abutment screws to 10 Ncm. Fit of the implant framework castings was evaluated by measuring the marginal opening between the casting and implant at the reference points. These measurements were averaged and statistically compared for differences.The mean marginal openings at the most distant measuring locations from the tightened retaining screw at location 1 was between 44 to 48 microm for the centrifugal system compared to between 28 to 32 mm for KDF (P < .01). For screws tightened at location 3, the mean marginal openings at the most distant measuring locations were between 40 to 51 mm for the centrifugal system compared to between 27 to 29 microm for KDF (P < .01).In comparison with the centrifugal casting and oxygen-propane system, the argon vacuum system was more accurate and user friendly and less technique-sensitive.The argon vacuum casting machine tested produced more accurate, better fitting implant-supported prosthesis frameworks than a conventional centrifugal casting system. The "1-screw" method of evaluating casting fit was most effective when either of the prostheses' end screws were tightened.