Objectives: Multislice computed tomography and cone beam computed tomography scans data are used at a variety of applications in medical and dental fields. Among these applications is image registration that has become an essential step in medical images processing as radiation therapy, oral and maxillofacial surgery, orthodontics, …etc,. The aim of this study was to evaluate the accuracy of MSCT- and CBCT- based registration for oral and maxillofacial surgery on dry human mandibles.Materials and Methods: The current study comprised ten dry human mandibles scanned twice using a MSCT scanner and scanned twice using a CBCT scanner. The registration accuracy was evaluated by superimposing both images sets for each mandible (of the same modality) then measuring the distance between the two images at definite points. This was done with both modalities. To compare various point based registration techniques, the two images of the same mandible were superimposed three times with different registration points; bone landmarks, dental landmarks, and fiducial markers then the distance between the two images was measured for the three techniques. This was done with both modalities.Results: The results of the current study demonstrated that CBCT showed lower mean discrepancy in the measurements than MSCT.Conclusions: The process of computed tomography-based registration expresses some sort of sub-millimetric inaccuracy that should be put into consideration according to the clinical needs. CBCT-based registration showed more accurate results than that of MSCT. Very good intra-observer agreement reflects the reliability of CBCT-based registration in follow-up process for different clinical applications in maxillofacial region. Markers-based registration technique was the most accurate while dental landmarks-based registration technique was the least accurate.