In this paper adaptive control of a brushless DC motor (BLDCM) using neural network identification and pole shifting (PS) controller is presented. Proper system identification is one of the important factors that gives a good controller performance. This means that when the model parameter estimates are good, the controller output is good, whereas if the model parameter estimates are bad then almost surely the computed control will be bad. Proper selection of the identified system model order is also investigated. A comparison study between fuzzy logic controller and the proposed controller is also investigated.