In this paper, we study the characterization of admissible linear estimators of regression coefficients in the general growth curve model with respect to an incomplete ellipsoidal restriction under the loss function (d(Y)-KBL)′(d(Y)-KBL). By the methods of linear algebra and matrix theory, when KBL is estimable, the necessary and sufficient conditions for linear estimators to be admissible in the homogeneous and non-homogeneous linear classes are given. If KBL is inestimable, we obtain the necessary and sufficient conditions for linear estimators to be admissible in the homogeneous and non-homogeneous linear classes using orthogonal projection approach.