An industrial standard steel ball plate of dimension 420 mm with 25 ceramic balls in a 5 × 5 arrangement was circulated among 12 European NMIs. The measurement task was the calibration of the centres of the 25 balls. The measurements were conducted between 2004 and 2006. One NMI withdrew from the comparison. The other 11 NMIs measured the ball plate and delivered data. The artifact was found to be stable during the duration of the comparison. For the analysis both the deviations from the KCRV and the En numbers were calculated. Most results agreed very well within the claimed uncertainty. A few results showed some larger deviations, e.g. length periodic modulations of the differences from the KCRV which still were within the claimed uncertainty, but might be a basis for some optimization of the measurement process. However, because of the significant length of time between the measurements and the final report, the optimization was already completed in most cases.