Atrazine is the most frequently detected herbicide in groundwater. For these reasons, atrazine's dynamics need to be studied, and work is needed to develop and calibrate mathematical models to help in determining best management practices to minimize contamination. This study was designed to make a small contribution in solving some of the environmental related problems;In the first part of this study, macroporosity of an agricultural soil in Iowa was characterized. This study confirmed that under saturated and partially saturated flow conditions, macropore flow was a dominant mechanism for water and chemical transport;Atrazine's movement through the soil profile was investigated in two, but representative, soils of the Sorraia Valley region, in Portugal, under irrigated conditions. These soils (alluvial and sandy soils) were monitored for atrazine and soil water content during the summers of 1996 and 1997. The alluvial soil received surface irrigation during both summers, and atrazine was surface applied at a rate of 1.2 kg a.i./ha. The sandy soil received surface irrigation in 1996 and sprinkler irrigation during 1997, where atrazine was applied at rate of 1.0 kg a.i./ha;The Root Zone Water Quality Model (RZWQM) is a one-dimensional model that simulates the biological, physical and chemical processes in the root zone. The pesticide component of RZWQM was calibrated and evaluated for two distinct conditions: drained soils in Iowa, and irrigated soils in Portugal. To calibrate and evaluate this model for Iowa conditions, 3 years of field data from a water quality research site at Nashua, Iowa, were used. This study showed that the range of predicted atrazine concentrations was within the same order of magnitude as observed concentrations;The data collected in Portugal was used as well to evaluate the model for alluvial and sandy soils, under Portuguese conditions. This study showed that RZWQM can be used to simulate soil water content and atrazine concentrations in the soil profile, under two different hydrologic conditions;Computer models should be easy to use, however, most of the available models are not very user friendly, making its use difficult and not appealing to the user. An interface between RZWQM and ArcView GIS was developed to simplify the data input in RZWQM, account for spatial variability in the model, and enable the user to visualize model output.