Three-dimensional measurements of surface topography were performed using a confocal laser scanner and a contact stylus instrument. Three surfaces known to be difficult to evaluate were chosen to be measured on the same area with the two instruments. The measurements from the optical and the contact stylus profilometer were compared with each other and with measurements obtained from high-resolution atomic force microscopy, which served as a reference instrument. Six implants manufactured from commonly used biomaterials were also measured on the same part of the implant, but not on the same area, with the optical and the contact profilometer in order to simulate the measurements that would be performed when different laboratories measure similarly treated surfaces. The numerical and visual differences achieved when measuring the same area with the two instruments investigated were compared. In general, we found an underestimation of the surface features with the contact stylus measurement and an overestimation with the confocal scanner. The stylus readings are mainly influenced by the radius of the stylus tip, the pressure of the stylus tip on the surface, and the hardness of the material. The optical profilometer has a tendency for creating spikes when surfaces with deep slopes are measured. For relatively soft metallic biomaterials, we found that using the optical instrument is the most appropriate method for surface roughness characterization, particularly when screw-shaped implants are analysed, whereas the stylus is preferred when larger areas with substantial slopes within the surface structure are to be evaluated.