The cytokinesis-block micronucleus assay in peripheral blood lymphocytes is one of the best standardized and validated techniques for individual radiation dose assessment. This method has been proposed as an alternative to the dicentric chromosome assay, which is considered the "gold standard" in biological dosimetry because it requires less time and cytogenetic expertise. Nevertheless, for application as a biodosimetry tool in large-scale nuclear or radiological accidents, the manually performed cytokinesis-block micronucleus assay needs further strategies (e.g., the automation of micronucleus scoring) to speed up the analysis. An essential prerequisite for radiation dose assessment is to establish a dose-effect curve. In this study, blood samples of one healthy subject were irradiated with seven increasing doses of x-ray (240 kVp, 1 Gy min⁻¹) ranging from 0.25-4.0 Gy to generate calibration curves based on manual as well as on automated scoring mode. The quality of the calibration curves was evaluated by determination of the dose prediction accuracy after the analysis of 10 blood samples from the same donor exposed to unknown radiation doses. The micronucleus frequencies in binucleated cells were scored manually as well as automatically and were used to assess the absorbed radiation doses with reference to the respective calibration curve. The accuracy of the dose assessment based on manual and automatic scoring mode was compared.