Robust statistics can be used to find estimates of true values and precision that are insulated from the effect of outliers. In this paper these procedures have been extended to inter-laboratory trials.