This study analyses records from 189 optometric practices in England and Wales on the glaucoma screening of 123,415 patients aged > 40 years during a prospective 6 month period. Its purpose is to compare the effectiveness of the various modes of screening which were used. All optometrists tested every patient by ophthalmoscopy. The 146 who, in addition, conducted tonometry on a routine basis detected a confirmed glaucoma in 0.35% of their patients, whilst the 43 who used tonometry 'on suspicion' detected a case in 0.25%, (P < 0.02). The 47 optometrists who conducted perimetry frequently (i.e. in > or = 15% of sight tests) detected a case in 0.46% of patients, whilst the other 142, who did little or no perimetry, detected a case in 0.29%, (P < 0.001). Eleven optometrists who relied mainly on ophthalmoscopy had a detection rate of only 0.12%. Optometrists with the most comprehensive modes of screening had the greatest referral accuracy. It was concluded that more widespread adoption of routine tonometry for people aged approximately greater than 40 years is necessary to reduce the present substantial number of false negatives; and that the frequent use of visual field analysis is additionally required to achieve the best results.