The article presents a modification of the Heliosat Method which has been proven to give a better estimation of solar radiation. In the modified algorithm, the absence of a cloud above each pixel is checked from its cloud cover index value, and if no cloud is detected, the global radiation incident on the pixel is computed using a clear sky model. Otherwise, the quantity is calculated following the original algorithm. The improvement obtained from the modification has been evaluated at a Heliosat Secondary User Station in Italy. The evaluation was done by comparing the hourly global irradiation estimated by the original method and that determined by the modified method with the values measured directly at 13 ground stations. The improvements of the global root mean square error and mean bias error computed from the combined data of all the 13 stations are 5.17% and 8.03% of the mean ground measured hourly irradiation, respectively. The improvements for the individual stations are as reported in Table 1. Based on the results of the evaluation, it has been concluded that the proposed modification gives a better estimation of the energy than the original algorithm. The improvement is more pronounced especially when the method is used in countries or regions where cloudless atmospheric conditions prevail for relatively longer periods of the year.