Transaction curtailment has become the chief method of relieving congestion to maintain system security and reliability. In the NERC curtailment model, all parameters are definite. However in practice, the power transfer distribution factors (PTDFs) are inaccurate and the limits to line flows are soft. In view of the fuzzy characteristics of the curtailment problem, the fuzzy curtailment set is defined first. Based on this curtailment set, the fuzzy constraints and objective are described and the fuzzy optimization curtailment model is presented. Second, using the maximum entropy method, this fuzzy optimization problem is solved as an unconstrained optimization one whose calculating procedure is both fast and simple. Finally, test results show that the fuzzy curtailment set can reduce the impact of the computational errors of the PTDFs on the transaction curtailment while the fuzzy optimization model will lead to less curtailment than does the crisp optimization model when slight violation of the operational limits is allowed.