The constraining force is one of the most important factors for a sustainable and economical choice of reinforcement in reinforced concrete slabs. Therefore, it is necessary to know the constraining force in the slab in case of a combination of vertical load and temperature. The objective of this research is to identify, which input parameters influence the constraining force. For this purpose 2D FEM models with an automated analysis process generate the input parameters for a sensitivity analysis. A Latin Hypercube Sampling with restrictions of the sampling region creates the calculated models. The Gaussian Process Regression generates surrogate models with \(R^2\approx 0.96 \ldots 0.99\) to easily compute more output values of constraining forces. The main and total sensitivity indices of the input values are derived by a variance based sensitivity analysis. The effective depth of the cross section, the tensile reinforcement ratio and the bending moment are the values with the largest impact on the constraining force. The upper and lower concrete cover, the length of the slab, the tensile strength of the concrete and the compressive reinforcement ratio have a low impact on the constraining force.