This prospective randomized study compared the immediate postoperative periglenoid radiolucencies among 3 glenoid-drying techniques used in total shoulder arthroplasty. Seventy-one consecutive patients with primary osteoarthritis underwent total shoulder arthroplasty by use of 1 prosthetic system with convex-back, keeled, polyethylene glenoid components; the same modern, instrumented pressurization technique was used to cement all glenoids. Of the shoulders, 21 had glenoid implants cemented after bony preparation with thrombin-soaked gel foam, 24 after compressed gas lavage, and 26 after saline solution lavage with sponge drying. The immediate postoperative anteroposterior radiographs were examined to evaluate the presence of periglenoid radiolucencies. Of the patients, 29 (41%) had radiolucencies evident immediately postoperatively, with all radiolucencies occurring in the faceplate zones. The mean total radiolucent line score was 0.63 (P = .94), with no significant difference among cementing preparation techniques (P = .89). Prosthetic mismatch did not differ among glenoid preparation techniques (P = .86). There was no statistical association between prosthetic mismatch and radiolucent line score either across (P = .62) or within (P = .99) the glenoid preparation groups. The associated costs in the gel foam group and compressed gas lavage group were 70 times higher than the cost in the saline solution lavage group. All radiolucencies were noted in the faceplate zones, with no radiolucency greater than 2 mm. Preparation of the glenoid surface for cementing showed no significant difference among the 3 techniques studied, although the material costs were significantly higher in the gel foam and compressed gas lavage groups compared with the saline solution lavage group.