Prolonged inspection of an adapting stimulus changes the appearance of a subsequent test stimulus. There are five distinct viewing conditions under which such 'after-effects' may be generated. These are MON-MON (inspect with one eye, test same eye), BIN-BIN (inspect with both eyes, test both eyes), BIN-MON (inspect with both eyes, test only one eye), MON-BIN (inspect with one eye, test with both) and TRANSFER (inspect with one eye, test with the other eye). A model based upon the assumption of the linearly additive effects of adaptation generated in 'dominance classes' of cortical units that are driven either by one eye, or the other eye, or by either eye or both eyes together, is described. This model generates predictions concerning the expected relative magnitudes of after-effects generated under the five viewing modes described above, and experiments are described that confirm these predictions. The model can be extended to generate predictions about other experimental conditions. A more complex version of the model is consistent with electrophysiologically derived estimates of the proportion of cortical units in each dominance class.