Our purpose was to quantify the Gilula score for measurement of lunate uncovering, to compare it with another method of measurement and to examine the reliability of these measurements in posteroanterior (PA) views in radial and ulnar deviation. Seventy-six normal wrist arthrograms were reviewed retrospectively. Carpal height and lunate uncovering measurements were made. Statistical analysis included mixed effects models to evaluate the difference between the mean measurements in each position. Reproducibility was assessed using imprecision estimates. Normal values for the Gilula method were 40% lunate uncovering in neutral, 49% in radial and 20% in ulnar deviation. There was a statistically significant difference between the values in the different views. Ulnar translation of the carpus can be measured reliably on neutral and radially deviated PA views using the Gilula method, but the different normal values for each view should be used. The Schuind method of measurement is comparable to the Gilula method in the neutral PA view.