This article presents numerical approximations to the asymptotic distributions of tests for structural change, specifically the Andrews and Andrews-Ploberger tests. These approximations, denoted as \( p^*(x) \), enable easy and accurate calculation of asymptotic p-values, which are crucial for applied economists. The author extends previous work by Hansen (1992) and MacKinnon (1994) by allowing the leading distribution to depend on an unknown parameter \( \eta \). The approximation method uses a weighted loss function over the p-value space, and the resulting approximations are highly accurate, even with a parsimonious model. The article also includes a GAUSS program for computing the test statistics and asymptotic p-values. The methodology is illustrated through empirical applications using autoregressive models, demonstrating how the approximations can provide more informative conclusions than critical values alone.This article presents numerical approximations to the asymptotic distributions of tests for structural change, specifically the Andrews and Andrews-Ploberger tests. These approximations, denoted as \( p^*(x) \), enable easy and accurate calculation of asymptotic p-values, which are crucial for applied economists. The author extends previous work by Hansen (1992) and MacKinnon (1994) by allowing the leading distribution to depend on an unknown parameter \( \eta \). The approximation method uses a weighted loss function over the p-value space, and the resulting approximations are highly accurate, even with a parsimonious model. The article also includes a GAUSS program for computing the test statistics and asymptotic p-values. The methodology is illustrated through empirical applications using autoregressive models, demonstrating how the approximations can provide more informative conclusions than critical values alone.