Glenoid loosening is one reason for failure of total shoulder arthroplasty. Several factors, including radiographic lucency, have been shown to be associated with glenoid loosening. The purpose of this study was to assess the correlation between glenoid design and immediate radiographic lucency in a prospective randomized clinical trial. Total shoulder arthroplasty was performed in 43 patients over a 2-year period. Twenty-three patients were randomized into the keel group and twenty patients into the pegged group. Postoperative radiographs obtained within 6 weeks of surgery were evaluated by 3 raters to determine glenoid lucency. On a scale from 0 (no lucency) to 5 (gross lucency and component loosening), the rate of lucency was 39% (9/23) in the keeled components, which was significantly higher than the rate of 5% (1/20) observed in the pegged components (P = .026). Patient age, gender, and glenoid size did not significantly affect glenoid component lucency (P > .05). The consistency reliability among raters (Cronbach alpha) was 0.87, and the intertester reliability was 0.87. Pegged glenoid components have less radiographic lucency when compared with keeled glenoid components in the immediate postoperative period.