This presentation shows the processes and methods for developing distributed rainfall-runoff modeling system using open source softwares. The objective of this study is to develop a MapWindow plug-in for running GRM (Grid based Rainfall-runoff Model) model (MW-GRM) in open source GIS software environment. MW-GRM consists of the GRM model, physically based rainfall-runoff model developed by Korea Institute of Civil Engineering and Building Technology (KICT), for runoff simulation, pre and post processing tools for temporal and spatial data processing, and auto-calibration process. Each component is integrated in the modeling software (MW-GRM), and can be run by selecting the MW-GRM menus. In developing MW-GRM, free software and open source softwares are used. GRM model was developed by using Visual Basic .NET included in Microsoft Visual Studio 2013 express, pre and post processing tools were developed by using MapWindow (Daniel, 2006) and GDAL (Geospatial Data Abstraction Library), and PEST (John, 2010) model was used in the auto-calibration process. The modeling system (MW-GRM) was developed as MapWindow plug-in. System environment was Window 7 64bit. MapWindow GIS ActiveX control and libraries were used to manipulate geographic data and set up GRM input parameters. ESRI ASCII and GeoTIFF raster data formats, supported by MapWindow and GDAL, were applied and shape file (ESRI, 1997) was used in vector data processing. GDAL is a library for translating vector and raster geospatial data. In this study, GDAL execution files were used to develop pre and post processing tools. The tools include data format conversion, spatial interpolation, clipping, and resampling functions for one or more raster layers. PEST is a model-independent parameter estimation software. Parameter estimation and uncertainty analysis can be carried out using PEST for model calibration and sensitive analysis. PEST is developed as an open source software, and single and parallel execution files are provided. This study developed GRM uncertainty analysis GUI as an interface system of GRM and PEST. GRM model had been a DLL type library including APIs to support developing another application. But PEST needs a model execution file, which can run in console execution window without user intervention. This study developed GRM execution file (GRMMP.exe) running in console window. It can simulate runoff using GRM project file, and no user intervention is allowed after the simulation has started. GRM uncertainty analysis GUI makes PEST input files (pcf, pif, ptf, rmf, etc.) by setting GRM parameters, observed data, PEST parameters, and selecting single or parallel PEST and PEST run automatically using GRMMP.exe file. In this study, all the functions necessary to develop GRM modeling system and pre and post processing tools could be implemented by using open source software. And MapWindow plug-in of GRM model can simulate runoff in open GIS environment including automatic model calibration using PEST. The study results can contribute to the wide spread of physically based rainfall-runoff modeling. And this study can present useful information in developing distributed runoff modeling system using open source software.