This thesis studies the effect of illiquidity on an investor’s optimal asset allocation, where illiquidity is the restriction an asset cannot be traded for an uncertain time period. The martingale method is used to derive the optimal asset allocation in three different cases: trading in the illiquid asset is always possible, trading in the illiquid asset is not possible at some points in time and it’s uncertain whether or not a trading opportunity in the illiquid asset arises. The cases are compared to measure the effect of illiquidity on the optimal asset allocation of an investor. It turns out illiquidity leads to a reduction of the allocation to the illiquid asset, especially for the less risk-averse investor. If we increase the number of time steps in the model, the investor will reduce his allocation to the illiquid asset if the sequence of no-trading points increases. This effect is even more persistent if we increase the evaluation period. However, the investor holds the same allocation if the the point in time the no-trading period arises changes.