Magnets for Nuclear Magnetic Resonance applications are designed to provide high level of magnetic flux density in a testing volume with the greatest level of field homogeneity. Anyway, the manufacturing and assembly tolerances cause unavoidable uniformity degradation, that is usually counteracted by correction coils. The measurement of the magnetic field is the final verification of the complex design and fabrication process of a magnetic system, aimed also at defining currents in correction coils to improve uniformity. Anyway, when seeking high accuracy, it is necessary to assess the effect on final uniformity of uncertainties in the measurement technique. In this paper, an approach is presented to assess impact of measurement error bars on final device performance, based on Worst Case Analysis, guaranteeing quick computational time.