Investigations of Type A (W-5 %Re/W-20 %Re) thermocouples were performed at several laboratories to validate their reference function before its standardization in the new issue of the international standard IEC 60584. The Type A thermocouples investigated were equipped with sealed protection tubes made of sapphire which were filled with an inert gas (argon). The investigations at Russian laboratories were performed mainly in carbon-free high-temperature furnaces. The calibration results obtained in the temperature range (600 to 1850) $$^{\circ }\hbox {C}$$ ∘ C in the carbon-free environment were within $$\pm 0.5$$ ± 0.5 % tolerance limits and confirmed the suitability of Type A thermocouples for industrial applications. In contrast, the Type A thermocouple 89/95-P investigated at PTB (Germany) was exposed to a carbon environment when annealed at $$1850~^{\circ }\hbox {C}$$ 1850 ∘ C and when it was calibrated at metal–carbon eutectic (Me–C) fixed points. Measurements made at Me–C fixed points did not deviate from the reference function by more than about 6 K at the first stage when temperatures were below $$1500~^{\circ }\hbox {C}$$ 1500 ∘ C . However, the inhomogeneity of the thermoelements increased continuously after the calibration at the Me–C eutectic fixed points. The additional measurements at the $$\hbox {Cr}_{3}\hbox {C}_{2}{-}\hbox {C}$$ Cr 3 C 2 - C peritectic fixed point ( $$1826~^{\circ }\hbox {C}$$ 1826 ∘ C ) resulted in a continuous emf drift to deviations from the reference function of about $$-$$ - (100 to 150) $$\upmu \hbox {V}$$ μ V which corresponds to a temperature equivalent of about $$-$$ - (9 to 14) K. The thermoelectric stability and homogeneity of the thermocouple 89/95-P during these investigations was checked by repeated measurements at the freezing point of copper ( $$1084.62~^{\circ }\hbox {C}$$ 1084.62 ∘ C ).