Orientation-contingent color aftereffects (CAEs) were studied using a color-cancellation technique for measurement. Procedures involved more than one period of inspection, each of which established CAEs, carried out successively to produce a combined or net CAE (akin to “nullification” used in other studies). When opposite color-orientation pairings were used in successive inspections, the net CAE was predicted faithfully by summation of the constituent CAEs as measured from each period of inspection independently, and thus showed qualitative as well as quantitative changes in coloration over time. This implies that “nullification” does not truly eliminate CAEs as has previously been assumed, and suggests that the units of measure used here may be linearly representative of CAE strengths. When the successive inspections used identical color-orientation pairings, however, summation was poor. This can be explained if inspection alters the mechanisms underlying CAEs, rendering retention of a successively established CAE of the same kind less effective.