Soil contamination due to spills or leaks of crude oils and refined hydrocarbons is a common problem. Estimation of spill volume is a crucial issue in order to determine the expected contaminating life span of contaminated soils. The direct procedure to determine the amount of hydrocarbon in soil is to measure the concentration of total petroleum hydrocarbon (TPH) in soil samples. The primary objective of this study was to assess the potential effects of oil contaminated soils on the water quality of Devegecidi dam reservoir. For this purpose, limited spill data available were evaluated and soil sampling studies were conducted in the Beykan oil field to analyze for TPH on oil contaminated soils. Available spill and measured soil TPH data were used in a subsequent modeling study to assess the reservoir water quality impacts due to dissolved mass leaching from hydrocarbon contaminated soils. Evaluation of available spill data between 1989 and 1995 revealed that a total of 252 recorded spills resulted in a net spill of 395 tons. The major types of oil spills were identified as well heads (WH), return lines/flow lines (RL/FL), and power oil lines (POL). A total of 211 soil samples was collected at selected well heads and analyzed for TPH in the laboratory. TPH results revealed a concentration range between 600 and 115 500 mg kg(-1) with a mean concentration of 20 300 mg kg(-1). Modeling studies focused on behavior assessment and involved two major components. The first component is a soil-leaching submodel for estimating the leachate concentration and contaminant mass leaching out of the contaminated soil body. The second component is a reservoir water quality submodel assuming complete-mix conditions for estimating the changes of hydrocarbon concentration in the reservoir water as a function of time. These two components are coupled via a mass inflow term present in the reservoir water quality model, accounting for contaminant mass loading contributed by the leaching of contaminated soil. Simulation runs performed under conservative conditions assuming an annual average oil spill volume of 95 tons and the minimum reservoir volume of 7.3 x 10(6) m(3) revealed that there is no imminent threat to reservoir water quality from the dissolved phase contaminants soils. Limited amount of available measurements of TPH concentrations in soil samples and benzene concentrations in reservoir water samples supported model results.