This study on the yearly means of some meteorological parameters in some localities of Chad solves the problem of lack of meteorological data which is very crucial in developing countries. The main idea is to treat by the least square method the existing chronological series to establish regression functions for further use in the problems of interpolation and extrapolation. Our decision is based on the comparison of the degree of fitnessε of these functions and their corresponding standard deviation s. When|ε|< s we conclude that the function is useable for interpolation and extrapolation in order to find missing meteorological data. For the interpolation we have obtained good results. What concerns the extrapolation, it should be done with care particularly for the relative humidity of the air because of its high variability and also when the interval of time in sufficiently long.