A method is proposed to characterize the sky and cloud conditions (SCC) by using ground based global/diffuse irradiance data. The term SCC is defined for the purpose of this paper as the bulk effect of clouds, and other atmospheric constituents on the values of the diffuse fraction of solar radiation (D/G), the ratio between the global to extra terrestrial radiation (K-T) and the transmissivity T-r measured at the earth＇s surface. However, there is a difficulty in using the measured values of K-T and D/G for describing the SCC since these values depend on the solar zenith angle and not only on meteorological factors. This difficulty is overcome by the use of a simple semi-empirical procedure by which one can eliminate the seasonal and latitudinal variations (due to variations in the solar zenith angle) in D/G and K-T, or similar relations, leaving only variations due to SCC. The procedure is based on a paper by Zangvil and Aviv on the effect of latitude and season on the relation between the diffuse fraction of solar radiation and the ratio of global to extraterrestrial radiation, Solar Energy 39, 321-327 (1987) where a numerical simulation of the seasonal and latitudinal (zenith angle dependent) changes of positions of points representing different simulated SCC on a K-T-D/G diagram was performed. The method is demonstrated on monthly mean D/G-K-T relations calculated from three years of global and diffuse irradiance measured at Sede Boker, Israel. Results show that the large scatter of the original data points on a D/G-K-T diagram is, to a considerable extent due to a seasonal effect (variations of the solar zenith angle). After these effects are removed by the use of the proposed technique the transformed monthly data points have much less scatter and they describe the observed seasonal variations of sky and cloud conditions correctly.