During general anesthesia, the eye position of patients with strabismus, measured by Krimsky's test at 1 m, 30 minutes after induction, has a linear correlation with the eye position measured clinically preoperatively. A similar study was performed on 77 patients; however, Hirschberg's test was used for simplicity and was performed as soon as the patient was in a surgical plane of anesthesia. We confirmed that there was a linear correlation. We also adjusted the surgery performed in seven cases where the test was anomalous, ie, when the eye position under anesthesia was more than 15 prism diopters more or less than expected based on preoperative clinical measurements. We did approximately 1 mm more or less surgery than we had planned on each eye muscle. This significantly improved the final results in anomalous cases. Patients with anomalous tests who did not have such adjustments had significantly poorer results.