Cumulative frequency distribution curves (CDC) for daily global irradiation on the horizontal produced by Liu and Jordan in 1963 have until recently been considered to have universal validity. Results obtained by Saunier et al. in 1987 and Ideriah and Suleman in 1989 for two tropical locations, Ibadan in Nigeria and Bangkok in Thailand, respectively, have thrown into question the universal validity of the Liu and Jordan generalized CDC. Saunier et al., in particular, showed that their results disagreed with the generalized CDC mainly because of differences in the values of the maximum clearness index (Kmax), as well as the underlying probability density functions. Consequently, they proposed two expressions for determining Kmax and probability densities in tropical locations. This paper presents the results of statistical analysis of daily global irradiation for Kumasi, Ghana, also a tropical location. The results show that the expressions of Saunier et al. provide a better description of the observations than the generalized CDC and, in particular, the empirical equation for Kmax may be valid for Kumasi. Furthermore, the results show that the values of the minimum clearness index (Kmin) for Kumasi are much higher than the generally accepted value of 0.05 for overcast sky conditions. A comparison of the results for Kumasi and Ibadan shows that there is satisfactory agreement when the values of Kmax and Kmin are comparable; in cases where there are discrepancies in the Kmax and Kmin values, the CDC also disagree.