This paper proposes a preliminary-test shrinkage estimator for the sca
le parameter in pooling data from extreme-value distributions. The opt
imum values of shrinkage coefficients for the preliminary-test shrinka
ge estimator are obtained based on a regret function. The optimum valu
es of shrinkage coefficients and the critical values for the prelimina
ry test are tabulated for several sample sizes. For a mean square erro
r criterion of goodness of estimation, the preliminary-test shrinkage
estimator is preferable to the usual preliminary-test estimator.