The aim of the work is to find the best solution for statistical sampling technique in validation of radiation sterilization dose (RSD) for biological tissues, according to ISO standard. As a model for sampling are biological tissues retrieved from one cadaver donor which consist of frozen bone grafts (18 packets), lyophilized allografts (68 packets) and demineralized bone powder grafts (40 packets). The size and type of products vary from long bones, cancellous chips to bone powders, tendons and facia lata, that make the number of bioburden per product could not be treated equally. Frozen samples could not be considered as the same production batch with lyophilized samples, because of different processing and irradiation temperature. The minimum number of uniformed samples needed for validation per production batch size, according to ISO 13409, is from 20 to 79 and 20 of them will be used for the test sample size, i.e. 10 for bio-burden determination and the remaining 10 for verification dose. Based on the number of uniformed grafts, statistical sampling can be carried out on lyophilized and demineralized bone grafts, but not on frozen bone grafts. Bioburden determinations were carried out and validated according to ISO 11737-1. Results of average bioburden determination (cfu/per packet), using sample item portion (SIP) = 1, are 5 cfu/packet for lyophilized bone grafts and 4 cfu/packet for demineralized bone powder grafts. Verification doses obtained were 2.40 kGy for lyophilized grafts and 2.24 kGy for demineralized bone powder grafts. The results of verification dose were accepted and the RSD of 25 kGy is substantiated It can be concluded that a statistical sampling technique can be applied if all the grafts produced in the same process such as lyophilized, demineralized as well as frozen are assumed to be in one production batch regardless of sample uniformity such as size, type and weight; for this ISO 13409 can be applied for the validation of RSD.