Abstract Data concerning calibration and type tests of an extremity system based on XD100 rings (Harshaw) are presented. Dosemeter used was a thermoluminescent chip inserted in a sealed pouch and was read out using a Harshaw 6600 reader. Aiming to reduce the lower detectable dose and to improve reproducibility of response, two different methods of analysis were carried out and compared. The first performed a background correction based on an average value obtained from the evaluation of a group of non-irradiated dosemeters and manually subtracted. The second method employed a glow curve deconvolution algorithm. The dosimetric system complied with all ISO/DIS 12794-1 draft report suggested tests, on which this work was based. This work obtained results that point to the applicability of those recommendations to Brazilian institutes and suggest their adoption by the national regulatory commission.