Unexpected diurnal discrepancies between high-quality spectroradiometers were observed during the 2000 Nordic Ozone Group Intercomparison campaign. The spectral ratios of the irradiances showed a diurnal variation of approximately 2-9%. This cannot be explained by the nonideal angular response of the instruments' input optics in one plane (cosine effect). Instead, by using a radiative transfer model, we show that differences in the angular response in four azimuth planes have the potential to bias the measured data by up to 4.4% (azimuth effect). Other relevant factors are also discussed and quantified and are shown to be significant when diurnal changes in radiation are explained by environmental factors, or when measured data are compared with model or satellite data. Again, intercomparison campaigns have the potential to reveal errors that would otherwise remain undetected.