Wind is the second largest renewable energy source after solar. It is one of the fastest growing sources of electricity in the world and currently of wind energy is installed in the United States and an additional is under construction (Office of Energy and Environment Affairs, 2011). For the growth of wind electricity, one of the most prominent environmental concerns relates to the death of birds, bats and other avian species resulting from collision with turbine blades. This thesis develops a model that provides the optimal strategy of turning the turbines off in a wind farm for certain periods to mitigate bird mortalities. We first create a single turbine optimization model for each hour on each day of a single month. We maximize the expected revenue generation and limit the expected bird mortalities to a certain level to solve for the dates and times for which the turbine should be turned off. The optimization problem is found to be part of common class of problems called Knapsack problems and through experiments we conclude that a linear programming (LP) relaxation of the problem provides a near-optimal solution. We extend the single-turbine model to a multiple-turbine model applicable to a wind farm. In this case, we solve for the percentage of wind turbines that should be turned off to limit the expected bird mortalities to a certain level. Finally, we carry out an uncertainty analysis and estimate probability distributions over the outcome of optimal strategy of turning the turbine off. We consider the Cape Wind project as a case study and limit the analysis to only one species of endangered birds called the common loon. We find that in order to save an expected number of 10 such birds in the month of March; we need to turn the turbine off for a total of 23 hours spread over specific dates and times. The average cost per bird was found to be $171.