Oculokinetic perimetry (OKP) has been developed to screen for glaucomatous field loss but has relatively poor sensitivity when compared with threshold perimetry. Forty-two eyes from 42 patients with glaucomatous field loss on Humphrey threshold perimetry and 32 normals performed hand-held OKP under controlled conditions of refraction and lighting. Those who passed the standard test had their OKP cutoff determined with increasing neutral density filters (NDFs) at a new point 15 degrees from fixation in the inferotemporal field (where first glaucomatous defects are rarest). OKP was then repeated with a NDF that increased the ambient light by 0.3 log units from cut off. Of the 16 glaucomatous eyes to pass standard OKP, 9 failed the NDF test, improving the sensitivity from 62% to 83% (p < 0.05). These 9 eyes had field defects that were significantly less severe (mean defect 4.85 vs 7.91 (p < 0.05) and corrected pattern standard deviation 4.12 vs 7.00 (p < 0.05) and were from younger patients (mean age 56 vs 66 years; p < 0.05) than those who failed standard OKP. None of the 32 normals failed standard OKP and only 1 of 32 failed the NDF test. The use of NDFs to customise OKP, producing essentially a staged suprathreshold contrast sensitivity test, appears to increase the sensitivity of the OKP screener without degrading its specificity, particularly in younger subjects.