Abstract In this thesis, we aim at investigating a model for option pricing to reduce the risks associated with agricultural commodity price fluctuations in Ethiopia. The daily closed agricultural commodity prices such as Washed Sidama class A Grade 3 (WSDA3) coffee price, Unwashed Nekempti grade 5 (ULK5) coffee price, Whitish Humera Gondar Sesame Seed Grade 3 (WHGS3) price and Whitish Wollega Sesame Seed Grade3 (WWSS3) prices, which are obtained from Ethiopia commodity exchange (ECX) market, are used to study the price movements. The nature of the log-return of these agricultural commodity prices exhibit heavy tails and high kurtosis. Jump diffusion models are used for modeling and option pricing of agricultural commodity prices. The method of maximum likelihood estimation (MLE) is employed to estimate the parameters for the models. The root mean square error(RMSE), the quantile-quantile (Q-Q) plot and the method of non parametric fit with normal kernel are used to test the validation of models. These tests indicate that the models fit very well to the observed agricultural commodity price data. We followed martingale approach option pricing strategy to reduce the risk caused by price fluctuation under risk neutral measure. Monte Carlo simulation is used to find the option prices of agricultural commodity price. The small changes in each parameter value are considered to test how much each parameter is sensitive to the models. It is observed that the jump component parameters show great effect on option prices. Different values are also taken for risk free interest rate to evaluate option prices of commodity price. Thus, an increment in the interest rate will result in raising the option prices and a decrement in the interest rate will yield in falling the option prices. Therefore, from the empirical results, we conclude that the models perform efficiently for modeling and option pricing of commodity prices to reduce the risk associated with price