Visual evoked potentials by different check sizes in patients with multiple sclerosis.

Visual evoked potentials (VEPs) were obtained using three checks of different size (25', 50', and 100' of arc) in 20 normal controls and 1,006 patients. The test was abnormal in 270 patients regardless of the check size, whereas in 53 patients, a prolongation of P100 latency occurred only with checks of a certain size. The procedure was most sensitive in detecting abnormalities with combined application of 25' and 100' checks and least sensitive with the 50' checks alone. In occasional subjects showing W-shaped responses, the use of another check size often elicited the usual V-shaped pattern, alleviating the problem of selecting the peak for latency determination.