This study investigates the possibility of establishing monthly-specific equations for estimating global solar radiation across Sudan. Using data from 16 stations, stretching along sub-humid, semi-arid, arid and hyper-arid zones, the derived equations embrace the dependence of global solar radiation on a single or a group of parameters, namely latitude, altitude and ratio of actual-to-maximum possible bright sunshine hours. Five kinds of regression have been tried, namely linear, exponential, power, polynomial and multiple linear. Extensive statistical comparisons between these country wide equations have been performed in order to rank them. A general conclusion is that the addition of geographical parameters to the classical predictor variable, i.e., the fraction of sunshine duration, improves the estimation of global solar radiation. Furthermore, using the relative duration of sunshine as a single independent variable is workable only if the regression equation is expressed in a nonlinear form, basically a second order relation. Estimating global solar radiation by the adopted approaches have proved sufficiently reliable. The developed equations could, thus, be used to predict this parameter in areas with no measuring systems or in equipped areas where there are missing data, so that a geographical distribution of global solar radiation across the country may be obtained.