In economics, an outcome is cost-benefit efficient if the utilities of all the concerned individuals are taken into account in making the decision leading to the outcome.

Example -

The mayor decides to demolish a house in the suburb to extend his own drive way. The benefit to the mayor is marginal, but the cost to the family that is put out of doors is much greater. Therefore this outcome is NOT cost-benefit efficient

Now, the mayor decides to demolish the house in the suburb inorder to allow a bypass that would save millions of drivers a lot time in getting to work. The total utility of the benefit the drivers would get is much greater than that of the cost bore by the family (or at least we assume so). Therefore, this outcome IS cost-benefit efficient.

Note: The median voter theorem is NOT cost-benefit efficient. To illustrate, consider the following:
  1. There are three voters (denoted A, B, C) each with $100 dollars, and each would vote for an issue which is to his/her benefit and against one that is to his/her detriment.
  2. An issue is brought up: Divide C's $100 and give $20 to A give $20 to B and burn the other $60.
  3. A will vote 'Yes', B will vote 'Yes', C will vote 'No'. Therefore the motion will be passed.

But note the outcome is not cost-benefit efficient. The gains to A and B is $40, but the loss to C is $100.