In the framework of a comparison between three laboratories, electron paramagnetic resonance (EPR) signal-to-dose response curves were measured for sets of 30 tooth enamel samples and the variance of EPR measurements in dependence on absorbed dose was evaluated, in nine combinations of laboratory of sample preparation and EPR evaluation, respectively. As a test for benchmarking of EPR evaluation, the parameters ‘critical dose’ and ‘limit of detection’ were proposed as performance parameters following definitions from chemical-metrology, and a model function was suggested for analytical formulation of the dependence of the variance of EPR measurement on absorbed dose. First estimates of limits of detection by weighted and unweighted fitting resulted in the range 101–552 and 67–561 mGy, respectively, and were generally larger with weighted than with unweighted fitting. Indication was found for the influence of methodology of sample preparation and applied EPR measurement parameters on performance of EPR dosimetry with tooth enamel.