This paper aims to evaluate the effect of splinting during implant impression. A master model with two fixtures at the sites of 45 and 47 was used. 20 impressions were made for all four techniques: (A) indirect; (B) direct, unsplinted; (C) direct, splinted; and (D) direct, splinted, sectioned, and re-splinted. Splinting was undertaken with autopolymerizing acrylic resin (AAR). Horizontal distance between fixtures was compared using a digital caliper. The difference in distance were analysed with one-way ANOVA. Group A showed a significantly lowest accuracy among all techniques (p < or = 0.05). There was no significant difference of accuracy among the groups using direct techniques (p > or = 0.05). Group D was more accurate compared to group B and C. We conclude that splinting of impression copings would be beneficial to obtain an accurate impression.