Over the last six years, CEN TC 127 and more in particular its working groups ad hoc 14 and ad hoc 7 have been active in evaluating and improving the reproducibility of fire resistance testing. Ad hoc 14 developed a draft procedure for the evaluation of the performance of fire resistance furnaces in 1994. The two working groups were subsequently responsible for an extensive series of proving tests aimed at evaluating the procedure, but also at investigating the possible role of the Plate Thermometer in harmonising thermal exposure. The results of the work are that, when furnaces are controlled using small diameter sheathed thermocouples, the thermal exposure of a test specimen differs greatly over the various furnaces. Controlling the furnaces on Plate Thermometer readings brings thermal exposures much closer together. This paper gives an overview and a quantitative explanation of the findings.