BACKGROUND In order to assess the influence of optical factors on the multifocal visual evoked potential (mfVEP), we obtained mfVEPs with optimal refraction and compared them to recordings with various degrees of dioptrical defocus. METHODS Monocular mfVEPs were recorded from the right eye in eight normal subjects. Dartboard stimuli with 60 sectors arranged in six concentric annuli spanning 60 degrees were generated with a VERIS system and presented on a computer monitor. Two pairs of electrodes were placed 3 cm above and below and 3 cm to the right and left of the inion. Two sets of mfVEP records per subject were obtained, one with best-corrected visual acuity and another when the stimulus was defocused by +1.0, +2.0 or +3.0 D. A signal-to-noise ratio (SNR) measure was calculated for every response from the two channels. RESULTS The effect of defocus depended on eccentricity: when defocus was at +2.0 D and higher, reducing visual acuity to <0.3, the central mfVEP responses were reduced to approximately 60%, while defocus had no marked effect at eccentricities >7 degrees. CONCLUSIONS The results suggest that, in contrast to the mfERG, the mfVEP requires optimal refraction to correctly assess the cortical responses.