Levelt's model is based on the assumption that the mean dominance time (t) of one eye is a function only of the stimulus strength (λ) in the other eye. He concluded that the assumption was supported by his experiments which were the two-choice rivalry task. But under the two-choice alternation process the assumption can also be expressed by another way that the mean suppression time (τ) of one eye is a function only of λ in the same eye. Any experiments under the two-choice alternation process can not test which assumption is correct. The present experiment, whose alternation process was intermittent, was performed to examine which assumption was correct. It has been proved that t is a function of λs in both eyes, but τ is a function only of λ in the same eye. It seems that Levelt's model should have been based on the latter assumption.