The effect of detergent type, pH and temperature during lysis on the DNA elution dose response was studied under non-winding conditions in exponentially growing, plateau-phase and synchronous S-phase CHO cells. Lysis with sodium-N-laurylsarcosine (NLS) increased the DNA elution rate and resulted in higher DNA elution for the same absorbed radiation dose than lysis with sodium dodecyl sulphate (SDS). This increase in elution caused a reduction in the shoulder width of the DNA elution dose-response curve, but did not significantly affect the final slope. One hour incubation at elevated temperatures (60 degrees C) during lysis either with NLS or SDS further increased DNA elution. Under these conditions DNA elution dose-response curves with a small or zero shoulder were obtained with exponentially growing, plateau-phase or synchronized S-phase cells. DNA elution was reduced to about 50 per cent of the controls when the pH of the SDS lysis solution was adjusted from 9.6 to 7.6. This effect was observed in cells that were lysed at room temperature, as well as in cells lysed at 60 degrees C. When NLS was used for lysis, a similar reduction in pH did not alter the DNA elution dose-response curve at either lysis temperature. Based on these results it is suggested that the shoulder observed in the DNA elution dose-response curve reflects partial separation of DNA from associated proteins. A direct and unconditional correlation of the DNA filter elution behaviour, as observed under non-unwinding conditions, with the induction of DNA dsb may thus not always be justified. Caution is required when elution results are used to establish correlations between the level of induction of DNA dsb and cell killing.