We present a calibration method for a full-Stokes polarimeter. The polarimeter uses two liquid-crystal variable retarders (LCVR) and a linear polarizer to measure the four Stokes parameters. The calibration method proposed in this paper calculates the errors in the experimental setup by fitting the experimental intensity measurements for a set of calibration samples to a theoretical polarimeter with errors. The errors calculated in the method include the axes alignment errors and the errors in the retardance values of both LCVRs. The resulting calibration parameters are verified by measuring the polarization state of a light beam passing through a rotating linear polarizer, a half-wave plate, and a quarter-wave plate and comparing with the predictions for an ideal, error-free polarimeter. It is found that an average reduction in rms error of 55.8% can be obtained with the proposed method.