D variant women occasionally form anti-D during or following pregnancy with a D-positive fetus. It is not known whether Rh immunization could be suppressed by using anti-Rh immunoglobulin (Rh Ig), or whether the injected antibodies would be absorbed by the woman's D variant cells. In order to predict the likely outcome, three anti-Rh Ig preparations were absorbed independently with five examples of D variant red cells: R1VIr (n = 3), R1Br (n = 2), and with cells of common Rh-positive and Rh-negative phenotypes: R0, r'r, r"r or rr cells (n = 1 each). The titres of the Ig preparations against all these cells were compared after three, six, nine and 12 absorptions. The titres of the unabsorbed anti-Rh Igs against R0 cells ranged from 2(15) to 2(18). Anti-D activities were hardly affected by absorption with r'r, r"r or rr cells, but were completely exhausted after three to six absorptions with R0 cells. Most of the D variant cells were as ineffective as rr cells in reducing anti-D activity against R0 cells, but one R1VIr variant produced a fall in titre against R0 cells, equivalent to four or five doubling dilutions after three to six absorptions, which could not be reduced further by subsequent absorptions. Therefore, a substantial proportion of the anti-D in the immunoglobulin preparation does not bind to D variant red cells. The component of the anti-D that does not bind to D variant cells would be expected to be present in the circulation of D variant individuals injected with anti-Rh immunoglobulin, and should be effective in suppressing Rh immunization. An increased dose of the anti-Rh Ig might have to be administered in order to allow for the fraction that can bind to D variant red cells and which would not be available for immunosuppression.