The aim of this in vitro study was to evaluate the effect of the type and design of the impression copings on the accuracy of implant impressions in 2 different conditions.A reference model with 2 implants inserted in bilateral mandibular canines was fabricated. The posterior teeth were inserted as tilted to simulate intra-oral undercuts. The teeth were eliminated to create an edentulous condition. Three different impression techniques were performed (open high retentive [OH], open low retentive [OL], closed [C]) for each condition. Total of 48 casts were made. Two scan-bodies were secured on each cast, scanned by an optical scanner. Then, they were compared to the scan of the reference model, and the calculated mean errors were analyzed with a 2-way ANOVA and Tukey test.There was no significant difference between the complete and partially edentulous groups (F = 3.252, P = 0.079). There was significant difference between the different designs of the impression copings (F = 31.789, P < 0.001) but the interaction between the dependent variables was not significant (F = 0.511, P = 0.603).The design of the impression copings was more important than the undercuts. The accuracy of the closed tray coping was greater than the low retentive coping and equal to the high retentive coping.