Pituitary tumors frequently compress the optic chiasm, causing visual field deficits. Surgical removal of these tumors could improve visual functions. Thus, predicting the prognosis of visual field function is required, but currently there is no method for predicting postoperative visual field status from preoperative data of tumor and optic chiasm characteristics. In this study, we performed preoperative evaluation of visual field prognosis using numerical parameters in 40 patients with pituitary tumors. Data from 30 patients were used to calculate the regression equation and those from the remaining 10 patients were used to confirm the validity of these equations. We defined quantitative values (area of tumor, A_tumor; curvature of chiasm, C_chiasm; and area of chiasm, A_chiasm) based on tumor size, tumor shape, and optic chiasm shape as determined using magnetic resonance imaging. We determined pre- and postoperative visual field sizes by ophthalmologic methods, and quantified them as numerical values (TNR). Postoperative recovery of the visual fields (obtained by comparing the post- and preoperative visual fields) was confirmed by increased postoperative TNRs (P < 0.01 for 4 isopters and P = 0.01 for 1 isopter, t-test). We attempted to predict postoperative visual field size using preoperative A_tumor, C_chiasm, and A_chiasm. Multiple regression analysis was performed, and three significant regression equations for predicting visual field size were obtained (N = 30, P < 0.01, F-test). The measured and predicted visual field sizes showed strong correlation (N = 10, r > 0.70). Thus, the quantitative parameters defined in this study clearly predicted postoperative visual functions in patients with pituitary tumor, and could find clinical applications in preoperative evaluation of visual field prognosis in neurosurgery.