Abstract: In this paper, La Motte’s results about admissibility in linear model are generalized. By using these results, the admissibility of quadratio and non-negative quadratio estimation of variance components in random one-way model is solved. It is also proved that the admissible class is a minimum complete class.