This thesis analyses the effectiveness of capital inflows controls in changing the size and composition of international financial flows. The major contribution is the focus on capital controls circumvention by the financial market and its limiting impacts on the effectiveness of short-run capital flows restrictions. At first, it is presented a theoretical approach to the question, discussing how it should be implemented. The model has a simple framework and considers two types of financial investors: patient and impatient. The first is more interested in long run returns, and the last gives more value for short run gains. We show that if the interest rate differential is endogenous to the controls fixation, and there is no possibility of regulations´ circumvention, then could be arranged a capital entrance tax scheme that induces the long run investment option for both investors, in absence of a confidence crisis. In occurrence of a higher probability of default than expected by the agent, the investment will be reverted unless higher interest rate differential is provided. Ex-post outflows capital controls could also avoid the investment reversion, nevertheless, we do not consider it here. If the interest rate differential is exogenous, which is typically the relevant case in Brazil, we might see an incentive to the investor to rename its capital as long run even when she expects to reverse the investment in the short run. We continue the analysis focusing the impact of the capacity of capital controls elision by the market. We incorporate in the model the circumvention of the entrance restrictions, showing that it ruins the capital controls effectiveness. In fact, the taxes of the capital controls are really determined by the market, and therefore, as more developed and sophisticated the financial market of the country is, more substitute assets and engineer financial operations can be used to circumvent the control. We underpin our conclusions describing several capital controls elisions operations in Brazil during the nineties, which make clear the difficulties of capital controls de facto implementation. We finish the dissertation with an impulse response function analyses, that indicates that the capital controls have only between two to six months of effectiveness in restricting financial capital inflows.