In this study, four impression procedures were assessed for accuracy in a laboratory model that simulated clinical practice. The accuracy of stone casts with brass implant analogs was measured against a standard framework. The fit of the framework on the casts was tested by manual and visual judgment and by microscopic measurement. The measurements supported the clinical judgments. When acrylic resin was used to splint transfer copings in an impression, all casts were acceptable and were more accurate than the best of the two other procedures, which did not use an acrylic resin splint. When an impression material was used to orient the transfer copings, the accuracy was better if undercut copings remained in the impression. When impressions were made by removing an impression from smooth transfer copings and replacing the copings in the impression, the majority of casts were unacceptable.