Purpose To identify the variable with the strongest association between the magnitude of the relative afferent pupillary defect (RAPD) and visual field indices in patients with glaucomatous optic neuropathy.

Methods Seventy-nine consecutive subjects with manifest glaucomatous optic neuropathy at least in one eye were enrolled in this retrospective study. RAPD was assessed with the swinging flashlight test and quantified with a neutral density filter. Perimetry was performed using the fast thresholding strategy German Adaptive Threshold Estimation. The values of the central differential luminance sensitivity (DLS), of the MD (mean defect) and of the ‘loss volume’ (LVOL) based on the individually modelled 3D hill of vision—the latter two within the eccentricities of 10°, 20° and 30°, respectively—were entered into a linear regression model without intercept as a function of RAPD.

Results An absolute value of RAPD of 0.3 log₁₀ units or more was present in 20 out of 79 glaucoma subjects (25%). The magnitude of RAPD was most closely associated with LVOL-30° (R²=0.77), followed by MD-30° (R²=0.73), MD-20° (R²=0.71), LVOL-20° (R²=0.67), MD-10° (R²=0.58), LVOL-10° (R²=0.54) and central DLS (R²=0.04).

Conclusions The prevalence of RAPD in glaucoma patients is comparatively small (25%). The magnitude of RAPD in glaucoma subjects is associated most closely with the LVOL within 30° eccentricity (which is the maximum visual field region tested in this study) and most loosely with central DLS, underscoring the impact of the entire (30°) visual field area on the afferent pupillary system.