The chromosome aberration yield for human lymphocytes exposed in vitro to various doses of 137Cesium has been studied. Dicentric, total acentric, and excess acentric data were seen to follow a Possion distribution. Calculated total hits demonstrated over-dispersion which could possibly be accounted for by a greater occurrence of single-hit phenomena being repaired than two-hit exchange processes. The resulting distribution generally contained an under-representation of cells with odd numbers of hits and an over-representation of zero- and even-hit classes as compared with Poisson predicted values. The relationship between dicentric yield and dose received in rads was fitted to the linear-quadratic formula Y = alpha D + beta D2 for dicentrics, yielding values of (20.1 +/- 3.8) X 10(-4) (aberrations/cell)/rad and (1.89 +/- 0.75) X 10(-6) (aberrations/cell)/rad2 for alpha and beta respectively. A plot of percent 'normal' cells versus the dose in rads resembled cell survival curves and was fitted to the relation P(D) = 100 e-Y where Y = alpha D + beta D2 with alpha = (23 +/- 11) X 10(-4) rad-1 and beta = (8.3 +/- 2.5) X 10(-6) rad-2. A possible use of scoring 'normal' cells for purposes of biological dosimetry is presented.