This article considers shrinkage preliminary test (PT) estimator for the ill-conditioned regression model under possible stochastic constraints. Four estimation strategies namely, shrinkage preliminary test ridge regression estimator (SPTRRE), shrinkage restricted ridge regression estimator (SRRE), restricted ridge regression estimator (RRRE) and unrestricted ridge regression estimators have been considered for estimating parameters. Using the quadratic loss function, the efficiency of SPTRRE relative to URRE are studied. It is observed that the SPTRRE dominates PTRRE uniformly and both RRRE and SRRE when the parameters moves away from the subspace of the restrictions. However, SPTRRE perform the worst followed by the RRRE and SRRE at the neighborhood of the subspace of the restrictions. It has been demonstrated numerically and analytically that SPTRRE provides wider range than the PTRRE which dominate the URRE. An optimal value for the size of PT for SPTRRE is given. It is noted that the size of PT for SPTRRE is more reasonable than the PTRRE for all ridge parameter k and ?