The pattern-reversal (P-VEPs) and the motion-onset (M-VEPs) of visual evoked potentials were modeled by means of three damped oscillators (O1, O2, O3) of identical construction. The O1, assumed to simulate the response of primary visual area (V1), was driven by the firing density of the lateral geniculate nuclei. O1 contributed mainly to the N75 and P100 peaks of the P-VEPs. The O2, driven by the O1 output, mimics the activity of V2, V3a, and MT. It contributed to the negative peak N145 of the P-VEPs or to the N160 in the M-VEPs. The O3 was suggested to model late slow processes probably of an attentive origin. The model parameters were set by optimization to follow the P-VEPs and M-VEPs obtained as a grand average of four young volunteers (PZ - A2 lead). The evoked potentials were described with normalized root mean square error lower than 13 %., J. Kremláček, M. Kuba, J. Holčík., and Obsahuje bibliografii