Objective: To evaluate the accuracy of robotic insertion of various ablation needles at various locations under computed tomography (CT) guidance in swine. Materials and methods: The robot was used for CT-guided insertion of four ablation needles, namely a single internally cooled radiofrequency ablation (RFA) needle (Cool-tip), a multi-tined expandable RFA needle (LeVeen), a cryoablation needle (IceRod), and an internally cooled microwave ablation needle (Emprint). One author remotely operated the robot with the operation interface in order to orient and insert the needles under CT guidance. Five insertions of each type of ablation needle towards 1.0-mm targets in the liver, kidney, lung, and hip muscle were attempted on the plane of an axial CT image in six swine. Accuracy of needle insertion was evaluated as the three-dimensional length between the target centre and needle tip. The accuracy of needle insertion was compared according to the type of needle used and the location using one-way analysis of variance. Results: The overall mean accuracy of all four needles in all four locations was 2.8 mm. The mean accuracy of insertion of the Cool-tip needle, LeVeen needle, IceRod needle, and Emprint needle was 2.8 mm, 3.1 mm, 2.5 mm, and 2.7 mm, respectively. The mean accuracy of insertion into the liver, kidney, lung, and hip muscle was 2.7 mm, 2.9 mm, 2.9 mm, and 2.5 mm, respectively. There was no significant difference in insertion accuracy among the needles (P =.38) or the locations (P =.53). Conclusion: Robotic insertion of various ablation needles under CT guidance was accurate regardless of type of needle or location in swine.