In a previous report (Finn et al., 1961) we gave reasons for thinking that the rapid removal of Rh-positive foetal erythrocytes from the circulation of a mother who was Rh-negative would prevent her from becoming immunized and producing Rh antibodies. We have now investigated the matter further, and the present paper describes the completed results of the earlier work (Experiment I) and then gives details of some subsequent observations (Experiments II and III). The reasoning involved and the scope of the investigations are first discussed. Levine (1943) demonstrated the marked degree of protection against Rh haemolytic disease afforded by ABO incompatibility between mother and foetus. We agreed with Race and Sanger (1950) that the probable mechanism lay in the rapid destruction of the incompatible foetal cells in the circulation by the anti-A and anti-B. Experiment I was therefore designed to determine whether immunization resulting from injection of ABO compatible Rh-positive blood into Rh-negative male volunteers could be prevented by the infusion of plasma containing high-titre saline-reacting (complete) anti-D antibodies, simulating the effect of anti-A and anti-B. Some initial results of this experiment have been briefly reported (Finn et al., 1962). It was found that, compared with control subjects, antibody formation was enhanced rather than prevented and we noted that in some cases nearly half the injected cells survived in the circulation for over a week, though at that time free complete anti-D could be detected in the recipients' sera.On the basis of these results a second approach which forms the substance of Experiment II was tried. Stern et al. (1961) Table I and to each of the 42 men in Experiment II in Table II. It was the aim in the second experiment to use 50 ml. of plasma with a high titre of incomplete antibody, thus attempting to achieve rapid clearance of injected blood, but because of practical difficulties this dose was not always possible.