In southeast Ohio, Raccoon Creek Watershed (RC) has an extensive mining history resulting in acid mine drainage (AMD) and subsequent environmental problems. Discharge measurements are collected by the United States Geological Survey (USGS) gage station at the Bolin Mills (BM) station. This data for the period 2011-2018 has been analyzed with the aim of identifying important parameters affecting the water discharge. Precipitation, antecedent precipitation index (API), and air temperature were the input variables considered in this study for transient data analysis using the program PAST, and modelling studies using Artificial Neural Networks (program NeuroShell). The result of the transient data analysis highlighted the quickflow, baseflow of watershed, delay time, and total response time. The maximum correlation coefficient (MCC) of BM discharge following API (0.22) has a stronger signal than the MCC of discharge following precipitation (0.14) while the cross spectral function of discharge following air temperature shows how air temperature strongly influences discharge, which is indicated from the high amplitude power (up to 65) produced from the spectrum. The result of the Neural Network Model with BM discharge as output variable and API, temperature, precipitation as input variables show underestimation of discharge, with the network unable to match modeled discharge with observed peak discharge. Comparison between network models of API using variable decay constants showed that an API of 0.74 produced the best results. Separate models using only precipitation were made to know the cause of this underestimation of discharge. The results were similar, with correlation coefficient of 0.748 for precipitation and 0.737 for API, proving that neither API nor precipitation contributed to this underestimation. The modeled data shows that the underestimation occurred mostly during winter to spring season for each year. Previous model of the discharge using the program BASINS (McKay,2017) suggests that this is because during the early spring, underground abandoned mines gets filled with water, resulting in a quick discharge from the mine. Based on this evidence, the model was divided into two section: high discharge during December through May and low discharge during June through November period. The result shows a better model match between the observed discharge and the modeled discharge with correlation coefficients higher than 0.9.