Dosimetry and biological effects of 40 and 50 keV low-energy X-rays generated by a SOFTEX Model CMBW-2 apparatus were studied. Doses were measured using a thin-window parallel-plate ionization chamber; beam quality was assessed using aluminum absorbers; exposure rates per unit current were determined according to the X-ray tube current and exposure times; and thermoluminescent (BeO chip) dosimeters were used to ascertain dose distributions in the irradiation field. The average correction factors for nonuniformity were calculated from the measured dose distributions. The means for ascertaining accurate exposures and doses using these methods are discussed. The dose-survival relationship of Chinese hamster V79 cells were assessed by irradiating them with 40 and 50 kV soft X-rays, 180 kV X-rays, and 60 Co gamma rays. Soft X-rays with three distinct effective energies were tested by changing the tube voltage kV and aluminium filter thicknesses; namely (1) 40 kV without filter, (2) 40 kV with a 0.2 mm thick aluminium filter and (3) 50 kV with a 0.7 mm thick aluminium filter. The effective energies obtained according to attenuation measurements using aluminium for these soft X-rays were 8.1, 11.7 and 18.5 kV, respectively. In this study the relative biological effectiveness (RBE) at 10 per cent survival compared with 60Co gamma rays ranged from 1.5 to 1.6. The RBE of 180 kV X-rays relative to 60Co gamma rays was 1.29. This study provided experimental data for the RBE of V79 cells in the intermediate energy range between hard and ultrasoft X-rays, data for which were previously reported by Goodhead and co-workers (1977, 1979, 1981).