To be useful in contact lens fitting, the videokeratoscope (VK) must be able to provide the practitioner with credible data on the vertex radius and central topography, or 'shape, of the patient's cornea. For this purpose, it is desirable that the measured value of the former should be credible at the level of 0.05 mm. In order to examine the accuracy, repeatability and reproducibility of the corneal dimensions provided by the videokeratoscope, eight eyes were measured on four different EyeSys systems at four independent sites. The order of measurement was random at each site. This amounted to a balanced uniform-level precision experiment ('ring test') of the VK instruments as defined in International Standard ISO 5725. The resulting data were analysed using the statistical procedures given in this Standard to provide a formal statement of the measurement precision of the EyeSys VK system. Based on this study, the estimated measurement precision of the EyeSys videokeratoscope as applied to the latter principal meridian of the normal human cornea is: Vertex radius: repeatability standard deviation, S(R)=0.0805, reproducibility standard deviation, S(R)=0.1041. p value: repeatability standard deviation, S(r)=0.0473, reproducibility standard deviation, S(R)=0.0574. The estimated measurement precision for the steeper principal meridian is: Vertex radius: repeatability standard deviation, S(r)=0.0771; reproducibility standard deviation, S(R)=0.1015. p value: repeatability standard deviation, S(R)=0.0698; reproducibility standard deviation, S(R)=0.0749.