Lymphocyte thymidine uptake in leucocytes prepared by NH4Cl induced RBC lysis was compared with that of lymphocytes obtained following Ficoll-Hypaque centrifugation using polynomial regression dose response curves which adjusted for inherent subject variability. Eleven male subjects were studied. Heparinized blood was either layered in Ficoll-Hypaque or treated with 0.87% NH4Cl to induce RBC lysis. Phytohemagglutinin-P (PHA) dose response curves were determined for each subject using triplicate 5 day cultures. For each subject, at each Pha concentration ranging from 12.5 to 167micrograms/ml, triplicate mean cpm of Ficoll-Hypaque treated lymphocytes and NH4Cl treated lymphocytes were compared by the 2 tailed Student's T test. Responses of cultures of NH4Cl treated lymphocytes were significantly different (less than 0.05) from the responses of cultures of Ficoll-Hypaque treated lymphocytes for at least 2 PHA doses in 8 of the 11 subjects. The quadratic dose response curves were also significantly different (p less than 0.05) for 9 of the 11 subjects. A repeated measures quadratic regression was used to determine the best fit dose response curves for each of the experimental conditions. For Ficoll-Hypaque treated lymphocytes, mean cmp=44,977+1400.4 (PHA concentration)-5.90 (PHA concentration)2+ subject effect; s.e.=22,256; for NH4Cl treated lymphocytes, mean cpm=58,039+1022.8(Pha concentration)-5.32 (PHA concentration)2+ subject effect; s.e.=37,966. The p values from the F tests for significance of these quadratic regressions are different, F=1.91, d.f.=30,30; p less than 0.001. The lack of significance of the NH4Cl treated lymphocytes' does response quadratic regression and its larger s.e. indicates that the NH4Cl treated lymphocyte culture dose response curve is less predictable than that of Ficoll-Hypaque treated lymphocytes.