In a clinical comparison of an XPERT NCT non-contact tonometer with Goldmann applanation tonometry (GAT), reported in the companion article, we concluded that the XPERT is a highly accurate and reliable instrument. By design, our evaluation differed significantly from earlier clinical studies in its use of multiple references (four calibrated GAT reference tonometers and seven GAT operators). Five clinicians were responsible for GAT measurement of 400 of a 620 eye population. Each clinician's population sample was essentially drawn at random from the common pool. XPERT vs. GAT performance indices were calculated for each of the five GAT operators. Because XPERT's measurements are essentially free of operator influence, it is reasonable to conclude that differences found among these five "XPERT performance" findings were primarily attributable to differences in GAT operator technique and endpoint judgment. Our findings disclose compelling evidence that GAT's accuracy and reliability depends on individual operator technique and experience. They indicate that in a clinical comparison, the protocol safeguards proposed by Wittenberg need be applied to ensure the integrity of studies using GAT as the reference tonometer.