OBJECTIVE To determine the reproducibility and the normal reference range of pulsatile ocular blood flow (POBF) values in healthy subjects using the Ocular Blood Flow Tonograph (OBF Laboratories, UK Ltd., Wiltshire, England). METHODS Pulsatile ocular blood flow was measured in one eye of each of 83 patients. Coefficient of reliability was determined by calculation of intraclass correlation coefficient via one-way analysis of variance. Mean difference between measurements was calculated for bias and first exposure effects. Pulsatile ocular blood flow from 163 healthy individuals were analyzed to determine the distribution, mean, standard deviation (SD), range, and the 5th and 95th percentile values. The influence of age, blood pressure, pulse rate, and intraocular pressure on pulsatile ocular blood flow was determined by regression analysis. RESULTS Reliability coefficient for pulsatile ocular blood flow values ranging from 290 microliters/min to 2,196 microliters/min was 0.92. Variation in bias and first exposure effect were not significant. Pulsatile ocular blood flow values were normally distributed. Mean values were 669.90 +/- 233.0 microliters/min in men and 841.90 +/- 254.6 microliters/min in women. Fifth and ninety-fifth percentile values were 364.75 microliters/min and 1,266.10 microliters/min in men and 397.18 microliters/min and 1,346.10 microliters/min in women. Pulsatile ocular blood flow was significantly influenced by pulse rate. CONCLUSIONS This study confirms the reliability of the Ocular Blood Flow Tonograph in repeated measurements of POBF within individuals over short time intervals. The high interindividual variation in POBF may invalidate comparison of POBF between individuals, and the wide range of normal values may limit the value of using a low POBF as a possible indicator of disease.