In solar engineering and atmospheric sciences solar irradiation and sunshine duration estimations are very important for different problems. Classical approaches that based on Angstrom equation for estimating global solar irradiation in terms of sunshine duration are abundant in the literature. Most of them include linear and lesser extent slightly nonlinear relationships between these two variables. Parameters are determined invariably by the least squares technique leading to regression lines or curves as models. None of these models provides within year variations in the parameters and they are all very rigid in applications yielding single global solar irradiation estimation for a given data set. In this paper, relation between extraterrestrial variables (length of day and solar irradiation) ratio and terrestrial variables (measured sunshine duration and solar irradiation) ratio is taken into account. This is a novel approach and has no restrictive assumptions. Decreasing amount of extraterrestrial variables ratio can be evaluated by subtracting measured terrestrial variables ratio (sunshine duration/total solar irradiation) from theoretical variables ratio. Differences between these ratios give atmospheric effects on length of day and extraterrestrial solar irradiation. The differences are represented by a parameter which helps to estimate easily the seasonal average parameter values without the application of the least square method. The main idea of this paper is to suggest not only a novel but also a practical formulation for solar irradiation and sunshine duration estimations. Its application is presented for three stations in Turkey and it is seen that better estimations are evaluated by proposed method than Angstrom method. Copyright (c) 2006 John Wiley & Sons, Ltd.