Groundwater contamination caused by agricultural fertilization is a widely recognized problem. In Hawaii, nitrogen fertilization from pineapple and sugarcane fields has posed a threat to several basal aquifers and has been implicated in coastal algae blooms. The concentration of nitrate-N in the Pearl Harbor basin on the island of Oahu was below 2.3 mg/L in the 1950’s and 1960’s, and has increased to as much as 7.6 mg/L in 1992 to 1994. The objective of this dissertation research is to develop a practical methodology for realistically estimating nitrate leaching from fertilized agricultural lands. Numerous mechanisms have impact on the distribution and migration of nitrate in the soil. Nitrogen fertilizer undergoes many N transformations and interactions with the soil and the plant after applications. In this study, an analysis of soil samples was performed to understand the leaching process of nitrate in the root zone of three different cropped fields in Hawaii. A detailed discussion is given to address various factors which control the nitrate transport process. To judge the sampling plan in relation to spatial variation, the field measurements were evaluated statistically by an uncertainty index, which is represented as the density of samples required for the estimate of sample mean of the nitrate concentration to fall within a defined limit of accuracy. In order to predict the effect of nitrogen fertilization on the groundwater contamination with very limited input data, a simple, analytical, lumped parameter model (LPM), was developed. The model can estimate the average nitrate leaching from the root zone in response to agricultural practices, N transformations and other related processes. The model was tested against the field data and two detailed numerical models, LEACHM-N and CERES-Maize. It provides an alternative way to assess nitrate leaching from the root zone with acceptable accuracy. A listing of the program is provided in Appendix 2. Owing to the complex nature of nitrogen behavior in the unsaturated zone, some degree of uncertainty is involved in the development of modeling approaches. In this study, five major sources of uncertainty were identified. These are: uncertainty due to spatial variation, uncertainty related to the accuracy of the input data, uncertainty due to simplifications in the development of the models, uncertainty due to the modeling parameters, and uncertainty due to the complexity of the unsaturated zone in Hawaii. The impact of these uncertainties on simulation results is evaluated. / Thesis (Ph. D.)--University of Hawaii at Manoa, 1996. / Includes bibliographical references (leaves 202-215). / UHM: Has both book and microform. / Water Resources Research Center, University of Hawaii at Manoa