The mean fluorescent intensity obtained using single antigen beads does not represent the titer of the serum sample very well because the degree of saturation of the antigen on the bead varies by individual bead and the human leukocyte antigen antibody-binding site on the bead may not duplicate the activity of the serum's antibodies in vivo. The flow crossmatch appears to come closer to correlating with the titer. As we have shown, the titer will roughly correlate with the flow crossmatch. This is consonant with the fact that the fluorescence level of the crossmatch bears a semi-quantitative relationship to the amount of antibody on the donor cell. The fact that successful transplants can be carried out when the median channel shift of the crossmatch is as high as twice the cutoff suggests that using the mean plus twice the standard deviation of the surrogate negative control may result in too many false positives.