The design of optimized histogram and probability graphs needed to detect a lognormal distribution in surveys of intraocular pressure (IOP), mean ophthalmic arterial pressure (OAP) and mean ocular perfusion pressure (OPP) is described. Applications are demonstrated in a sample of 58 normal adult eyes.A logncrmal distribution was found for IOP and for OAP and OPP estimates obtained with a noncon-tact noncompressive method. Conversely, the absence of an identifiable linear distribution for OAP and OPP estimates obtained with compression ophthalmodynamometry was noteworthy.The implications of lognormality for definitions of a normal range and for correlations involving IOP, OAP and OPP data are briefly discussed.