A comprehensive radiative transfer model is used to calculate surface spectral ultraviolet irradiance under cloud-free conditions. The results are compared with measurements made at Lauder, New Zealand (45¿S, 170¿E) before and after the eruption of Mount Pinatubo and including a snow-covered surface. There is reasonable agreement between measured and calculated diffuse, direct, and global irradiances over the range 300 to 450 nm. Discrepancies may be due to calibration errors in the UV measurements, or in the extra terrestrial irradiances input to the model. Ratios of diffuse/direct irradiances are independent of such calibration uncertainties and therefore provide a sensitive test of the model. If appropriate ozonesonde data, surface albedo, and aerosol optical properties are used, the model ratios are in satisfactory agreement with measurements over a wide range of observing conditions. For cases in which the atmospheric optical properties are best known the agreement is better than 8% in the UV-B range, and for wavelengths 320 to 450 nm the deviation is smaller. The comparison suggests that the ultraviolet radiation exposure can be computed with confidence for clear sky conditions if the appropriate atmospheric molecular density profiles, ozonesonde data, surface albedo, and aerosol optical properties are available. The UV radiation model is used to investigate the impact of changes in solar zenith angle, ozone abundance, surface albedo, and aerosol loading on UV radiation reaching the surface of the Earth. The ratios of diffuse to direct irradiance depend critically on solar zenith angle, surface albedo, and aerosol extinction. Ozone changes have pronounced effects on the global UVB irradiance but have only a minor effect on these ratios. ¿ American Geophysical Union 1994