Thirteen boiler steels were exposed in pure water vapour at 650°C for 32 weeks (5,376h). To assess their oxidation kinetics, four current discontinuous measurement methods were used. The results show that, for the T91 steel, the different investigated methods give similar results; for the other steels, the evaluation of corrosion rates depends on the method used. Thus their lifetime prediction is often over- or undervalued. This is mainly the case for steels with scales that spall off and/or steels with a non-uniform scale growth (localised oxide nodules for a rather long exposure time). This is particularly true in the steam environments which lead to oxidation processes with oxide scales different from those in dry air or oxygen. So the use of steam oxidation data and/or relations among them has to be done with caution because it can lead to a wrong prediction of performance.