The micronucleus assay has been considered as a fast method for determination of the chromosome aberration. It was shown in our earlier study that the micronucleus frequency was increased with the radiation doses in certain range, no matter what the radiation source is (x-rays, γ-rays and neutron). We proved firstly that micronucleus frequency induced by irradiation between ,in vivo, and ,in vitro, coincided well. That was the evidence for extrapolation value of the radiation dose in human from micronucleus graduated curve ,in vitro., Furthermore we got good result when it was used in a patient exposured accidently. This case was as follows:,Mr. Wang, male, was irradiated by ,60,Coγ-rays in an accident on 1st September, 1980. The frequency of micronuclei was found 87‰ 2.5 hrs after irradiation in cultured blood. The table showed the dose-response relationship from 4 healthy blood donors ,in vitro., The curve could be described by a linear-quadratic model Y=4.0271+ 15.3024 X (Y—micronucleus frequency, ‰: X—irradiation doses, Gy) with the dose range from 0.2 to 6 Gy, the correlation coeffecient r=0.994. From this model, we estimated that the irradiated dose received by this patient was 5.42 Gy. It was close to the dose from either physical calculation (5.22 Gy), or chromosome aberration, or other biological dosimeter. We consider the micronucleus assay can be used comletely as a biological dosimeter for patients exposured by higher dose. Although the sensitivity is less than that of chromosome aberration, it is less time-consuming, less expensive procedure and very useful to examine the large number of samples, especially in the nuclear accident.