Binocular interaction for a central field was studied with transient scalp visual evoked cortical potentials (VECPs) using two light-emitting-diodes. VECPs were obtained for binocular and monocular visions with dominant and non-dominant eyes, and arithmetical sums of monocular VECPs with dominant and non-dominant eyes were calculated. Amplitude and latency of remarkable initial three peaks were tested with the multivariate analysis of variance. Significant differences were noted among the four VECPs. Pairwise comparisons showed that (1) the amplitude of the first peak for the binocular VECPs was larger than that for the monocular VECPs but smaller than that for the sum-VECPs; the latency of the first peaks for the binocular VECPs were earlier than that for the monocular VECPs with the non-dominant eye; (2) the amplitude of the first negative peak for the sum-VECPs was larger than that for the binocular VECPs, and the peak latency for the sum-VECPs showed later than that for the binocular VECPs; (3) the amplitude of the second positive peak for the binocular VECPs and monocular VECPs with the dominant eye was larger than that with the non-dominant eye, but smaller for the binocular VECPS than that for the sum-VECPs; the latency for the binocular VECPs showed earlier than that for the monocular VECPs with the dominant eye and for the sum-VECPs. Binocular suppression was noted in amplitude for the three peaks and binocular facilitation was noted in latency for the latter two peaks.