The Robertson-Berger sunburn meter, model 500, has no temperature compensation, and the effect of temperature on the instrument response has been investigated and discussed in several reports. It is recommended to control the temperature of the detector or at least measure it. The temperature sensor is recommended to be positioned within the detector unit. We have measured the temperature at three different positions in the detector: At the edge of the green filter where the phosphor layer is placed; at the glass tube covering the cathode; and, finally, the air temperature inside the instrument. These measurements have been performed outdoors since July 1991, with corresponding measurements of the global and direct solar radiation. There was no difference between the temperature of the glasstube covering the cathode and the air inside the instrument, at any radiation level. However, there was a difference between the green filter and the two others. The difference is linearly dependent on the amount of global radiation. The temperature difference, (Delta) T (temperature between the green filter and the air inside the sensor), increased 0.8 degree(s)C when the global irradiation increased by 100 W/m2. At maximum global radiation in Trondheim (latitude 63.4 degree(s)N) (Delta) T was approximately 5 - 6 K when the global radiation was about 700 W/m2. This was valid for temperatures between 7 degree(s)C and 30 degree(s)C. Only clear days were evaluated.