Abstract: Experience and studies confirm that short duration rainfall of less than 2 or 3 h are of special significance for flood studies particularly with roads, highways and urban settlements in the high rainfall tropical regions of the world. The situation is compounded for Nigeria and some developing nations lying in the tropical zone where intense annual rainfall is high but scarcity of reliable hydrological data for planning and design of drainage works is scarce. For these peculiar situations, the need arises for development and use of statistical tools capable of extracting the maximum possible information content from data that are generally seen to be inadequate. In this study, rainfall Intensity-Duration-Frequency (IDF) studies were undertaken for Nigeria using annual extreme rainfall series available for 35 meteorological stations. In order to obtain floods of high return period with a good level of confidence that is needed in engineering practice, the whole country was grouped into 10 rainfall zones. The analysis was undertaken using the Extreme-Value Type 1 (EVI) distribution. For comparison purposes, the parameters of the (EVI) distribution were evaluated using the new concept of Maximum Entropy (ME) and compared with the results of the estimates obtained by the widely recommended Maximum Likelihood (ML) technique. The superiority of the new concept of Maximum Entropy (ME) for evaluation of the parameters of the EVI distribution is demonstrated in terms of its greater efficiency in the extraction of maximum possible information content from each group of data as well as its relative economic use of computer time. The IDF curves generated for different rainfall zones in Nigeria by the new method of Maximum Entropy (ME) are illustrated graphically.