Statement of problemAlthough the intraoral scanning of edentulous ridges is feasible, clinical evidence that the resulting denture retention is equivalent to that achieved with conventional impressions is lacking. PurposeThe purpose of this clinical study was to determine the retention of complete denture bases fabricated from digital intraoral scans versus conventional impressions by using border molding and posterior palatal seal compression. Material and methodsTwenty volunteers with an edentulous maxilla were recruited. An intraoral scan of the maxilla and a conventionally border-molded impression with a custom tray were made. The conventional impression was poured; the definitive cast was scanned. Three-dimensionally (3D) printed (PB1) and milled bases (MB1) were fabricated based on the scan of the definitive cast. Based on the intraoral scan, a 3D printed (PB2) and a milled base (MB2) were fabricated. On each base, a platform with a hook consisting of a central notch orienting the force against the post dam (PD) and 2 lateral notches orienting the forces against the left (LT) and right (RT) tuberosities was set in the center of the outer surface of the base. A traction dynamometer was inserted in the hook and oriented into the corresponding notch by applying force until dislodgement. All bases were subsequently stored in artificial saliva for 2 weeks and scanned. Retention testing was repeated by using the same procedure. To evaluate trueness and to visualize the differences on a color map, the scan of the definitive cast and the intraoral scans were matched and compared in 3 dimensions. The Wilcoxon tests were used to compare the retention of the different bases (95% confidence interval, α=.05). ResultsNineteen participants with a mean ±standard deviation age of 64.1 ±14.7 years completed the 4 study sessions. The retention of printed bases (PD: 16.08 ±15.28 N; LT: 14.98 ±14.72 N; RT: 11.28 ±9.57 N) and milled bases (PD:14.52 ±17.07 N; RT: 11.99 ±12.10 N; LT: 13.55 ±15.53 N) fabricated from conventional impressions presented significantly higher retentive forces than those printed (PD: 6.21 ±4.72 N; RT:5.12 ±2.78 N; LT: 4.45 ±2.77 N) and milled (PD: 6.58 ±4.92 N; RT: 4.65 ±2.63 N; LT: 5.02 ±3.58 N) from the intraoral scans (P<.05). The differences were significant in all directions of dislodgement, as well as after storage in artificial saliva for 2 weeks. Comparison of the 3D distances between the intraoral scan and the definitive cast revealed a mean deviation of 0.45 ±0.11 mm. ConclusionsConventional impressions of the edentulous maxilla, including the clinical steps of border molding and posterior palatal seal compression, provide better retention than digital intraoral scans with both milled and 3D printed denture bases.