State of the art methodology using experimental design and response surface techniques were used to perform uncertainty and risk analysis for managing reservoir uncertainties during history matching, production forecasting and production scheme optimization.In this paper we present the method applied successfully on a complex carbonate fractured reservoir with a long production history with approximately 500 producers and injectors. The simulation model is dual porosity/dual permeability, with 5 million cells. The objective is to quantify the impact of principle reservoir uncertainties on key field performance responses such as cumulative oil production, plateau length and water production using a risk analysis approach. The main uncertainties of the field were fracture connectivity, matrix-to fracture transfer, super-permeability streaks (Super-K) extent and matrix properties in a well spacing of one kilometer.From field performance experience, main parameters were identified and their responses during the historical production period were analyzed. The impact of parameters on history match quality was analyzed and an appropriate range of uncertainty was selected. A sensitivity analysis was performed to prioritize and rank the key influential parameters impacting production forecast. A novel methodology was applied to convert discontinuous realizations into a more continuous problem. Using the most influential selected parameters, another sensitivity analysis was done to accurately predict the dispersion in the production responses. A polynomial regression equation was derived for each of the main production responses. Finally, probabilistic distributions were generated for key production responses. This paper will highlight results of the study that helped in giving clear direction in field development planning by identifying and quantifying the main drivers influencing production behavior. Additionally, the results allow for quantifying the key uncertainty of the model forecast and associated risk to be investigated further. The results of the study can also be used for production optimization process by deriving the best operating constraints for highest recovery.