A routinely research of solar radiation is of vital requirement for surveys in agronomy, hydrology, ecology and sizing of the photovoltaic or thermal solar systems, solar architecture, molten salt power plant and supplying energy to natural processes like photosynthesis and estimates of their performances. However, measurement of global solar radiation is not available in most locations across in Nigeria. During the past 5 years in order to estimate global solar radiation on the horizontal surface on both daily and monthly mean daily basis, numerous empirical models have been developed for several locations in Nigeria. As a result, various input parameters have been utilized and different functional forms used. In this study aims at comparing, classifying and reviewing the empirical and soft computing models applied for estimating global solar radiation. The empirical models so far utilized were classified into eight main categories and presented based on the input parameters employed. The models were further reclassified into several main sub-classes and finally represented according to their developing year. On the whole, 145 empirical models and 42 functional forms, 8 artificial neural network models, 1 adaptive neural fuzzy inference system approach, and 1 Autoregressive Moving Average methods were recorded in literature for estimating global solar radiation in Nigeria. This review would provide solar-energy researchers in terms of identifying the input parameters and functional forms widely employed up until now as well as recognizing their importance for estimating global solar radiation using soft computing empirical models in several locations in Nigeria.