The aim of this paper is to extend iterative feedback tuning (IFT), which is a databased approach for controller tuning, with robustness constraints. Hereto a constrained IFT problem is formulated that is solved by introducing a penalty function. Essentially, the gradient estimates decompose into (a) the well-known IFT gradients and (b) the gradients with respect to this penalty function. Experimental results obtained from the motion control systems of an industrial wafer scanner confirm enhanced performance with guaranteed robustness properties