This paper investigates the finite sample behaviour of the Wald test of a slope coefficient (t-ratio) in sample selection models following the maximum likelihood estimation, specifically under multicollinearity identified by Nawata [Nawata, K., 1993. A note on the estimation of models with sample-selection biases. Economics Letters 42, 15–24]. The evidence shows that the conventional Wald test can perform very poorly under the multicollinearity problem, but the proposed bootstrap method can control the size successfully.