Manual hexadimethrine bromide (Polybrene) tests (Polybrene in low-ionic medium) were used in parallel with manual low-ionic-strength solution (LISS) procedures for the routine testing of patient samples referred to a general hospital blood bank. Of 5646 consecutive sera tested, 5167 (91.5%) did not react with either technique; 320 sera (5.7%) reacted in both methods. The Polybrene technique detected 63 antibodies which did not react in the LISS methods. One hundred sera did not react in the Polybrene test, but did react in the LISS methods. Sera showing discrepant results between the two methods were further tested in a reference laboratory. Polybrene tests appeared to be better in avoiding reactions due to clinically nonsignificant antibodies. The LISS methods, however, appeared to be more sensitive in detecting antibodies of potential clinical significance.