fmincon, nonlinear inequality constraints satisfied, but also minimized

I have fmincon running a nonlinear constrained problem, with inequality constraints. I have checked with the derivative checker that everything is fine. The optimization runs, but the constraints are being minimized even though they are below the boundary, and it makes the objective minimization go the wrong direction.

So why does it try to minimize well satisfied constraints, when it also worsens the objective?

[MiguelMauricio]

Hej Mads, could you show us a bit how is your problem? As you say, it should not try to minimize in the constraints direction. Have you checked that both the constraints and the objective function are properly defined?