Direct measurement of optical forces based on recording the change of momentum between the in- and outgoing light does not have specific requirements on particle size or shape, or on beam shape. Thus this approach overcomes many of the limitations of force measurements based on position measurements, which require frequent calibration. In this work we validate the achievable accuracy for direct force measurements in the axial direction for a single beam optical tweezers setup, based on numerical simulations and experimental investigations of situations, where the true force is known. We find that for typical experimental situations a good accuracy with an error of less than 1 % of the maximum force can be achieved, independent of particle size or refractive index, provided that the total amount of light scattered in the backward direction is also taken into account, which is easy to accomplish experimentally. Due to the inherent particle shape independence of the direct force measurement method, these findings support that it provides accurate results for 3D force measurements for particles of arbitrary shape.