Pooling of screening red cells used in pre-transfusion testing has been discouraged by many workers since it might lead to the failure to detect alloantibodies. However, there is minimal scientific data to support this view and in the UK a high proportion of blood transfusion laboratories were still using pooled red blood cells for antibody screening in 1982-1983. In order to solve the dilemma of sensitivity of serological tests when using pooled versus individual screening cells, a comparative evaluation of both approaches was carried out. One hundred and five sera known to contain weak, warm-reacting clinically significant alloantibodies were tested by the three same standard sensitive techniques using samples of individual screening cells and a pool of the same cells in parallel. Our results show that when pooled cells were used, 12% of the selected alloantibodies were undetectable and a further 23% were only detectable microscopically. All antibodies examined were easily detectable macroscopically when individual screening cells were used. It is concluded that the sensitivity of pre-transfusion screening tests is reduced when pooled red cells are employed as opposed to individual samples of screening cells and that the use of pooled cells should be discouraged in routine pre-transfusion and antenatal antibody screening.