Nearly half of the population of the United States lives in coastal regions, and millions of visitors from across the nation and world enjoy the coasts every year. Coastal and marine areas provide for recreation, economic activities essential for the financial health of the nation, and vital ecological services. As they provide so many benefits to the U.S., it is vital to protect and preserve the coastal and ocean areas from the increasing, competing demands they are facing. In order to protect and preserve these complex systems, a comprehensive approach incorporating science, engineering, humanities, and social sciences should be taken; this approach is commonly referred to as Ecosystem-Based Management. This dissertation focuses on developing a framework that can be used to identify appropriate sub-regions in Northern Gulf of Mexico coastal and marine environments for the purposes of Ecosystem-Based Management. Through this work, the roles of three management protocols used for managing coastal areas – coastal and marine spatial planning, ecosystem-based management, and integrated ecosystem assessment – were examined individually as well as their integrations with each other. Biological, ecological, physical, human, and economic indicators for partitioning an ecosystem were developed and weighted for each management protocol using the analytic hierarchy process and expert elicitation. Using the weighted indicators, a framework for identifying sub-regions and estuarine classification system was developed. The framework and classification system were applied to five estuaries within the Northern Gulf of Mexico: Barataria, Galveston, Mobile, and Perdido Bays and Mississippi Sound. Initial results from this work show that: 1. Sub-regions can be identified as associated to each other based upon indicator data values and not upon physical location. 2. Even though the weights calculated for the management protocols vary significantly, for systems that were not highly homogeneous in indicator data values, the different weights did not produce the vastly different cluster maps expected. 3. The scale work indicates that to identify appropriate sub-regions using the developed framework, a larger grid size produces more consistent results for larger systems whereas a smaller grid size produces more consistent results for smaller systems. Recommendations for further research are also presented.