Brain tumor detection is challenging task due to complex structure of human brain. MRI images generated from MRI scanners using strong magnetic fields and radio waves to form images of the body which helps for medical diagnosis. This paper segment the MRI image of brain tumor into two class first is tumor area while other is non tumor one. Here by using Fire Fly algorithm segmentation of tumor region can be done without any prior training with high accuracy. Proposed algorithm utilize median filter as well for removing the noise part of the image. Experiment was done on real image dataset. Results are compared with existing methods on various evaluation parameters and it was found that proposed algorithm is better than others.