Polysulphone film is used as a personal UV dosemeter in dermatological or epidemiological studies. The relative efficiency of this detector does not exactly match the action spectrum as proposed by the International Commission on Non-Ionizing Radiation Protection (ICNIRP) and to which the UV dose and exposure limits refer. Therefore, the calibration of the dosemeter depends on the spetrum. In the present paper the variation of the calibration factor for terrestrial solar UV spectra is analysed on the basis of a two year observation period at a site near Munich. Germany. A detailed error estimation is included. It is shown that the variation of the calibration factor within this class of spectra is the main contribution to the total uncertainty of the dose determination, which can be up to 40%. The shape of the spectrum of terrestrial solar UV radiation is mainly determined by the total ozone column and the solar elevation angle. It is shown how the calibration depends on these two parameters and how this additional information can help to reduce the measurement error to a residual uncertainty of 17%. Exposure studies of terrestrial solar UV radiation using polysulphone film as a dosemeter would gain in accuracy if total ozone column values at the study's site could be measured or taken from satellite or weather service data. The interpretation of the magnitude of the dose uncertainty depends on the further use of these data.