A focimeter is one of the basic clinical instruments used in every optometric practice, and verification of the accuracy and calibration of the instrument are of the utmost importance. Calibration is accomplished using special test lenses. These lenses must be of high quality and of nominal ophthalmic power that is known with high accuracy. The standard ISO 9342:1996 Optics and Optical Instruments: Test Lenses for Calibration of Focimeters imposes stringent requirements on permitted deviation from nominal ophthalmic power of test lenses. In this study, the influence of tolerancing and uncertainties in design and production of standard test lenses for calibration of focimeters was analyzed. Two approaches were used. First, the paraxial approximation was used to relate the lens parameters with ophthalmic power and to calculate the uncertainty budget. Second, exact ray tracing using a professional optical design program and tolerancing sensitivity analysis in conjunction with Monte Carlo simulation were used to evaluate the uncertainty budget and balance the required tolerances. The results of both approaches are compared and discussed in detail. The customary uncertainty budgeting based on simple paraxial approximation does not guarantee the compliance with the requirements of the ISO 9342:1996 standard. Consequently, such an approach is inadequate for characterization and tolerancing of test lenses for calibration of focimeters.