This paper presents computationally convenient approximations to the asymptotic p-value functions for the Andrews and Andrews-Ploberger structural-change tests. These approximations enable accurate calculation of asymptotic p-values for these tests, which have nonstandard distributions depending on two parameters: the number of parameters tested and the range of the sample examined for the break date. The paper introduces a parametric approximation method for the p-value function, using a polynomial function combined with a leading distribution function, such as the chi-squared distribution. This approach allows for accurate and efficient computation of p-values, even for small p-values, and is applicable in a wide range of nonstandard statistical contexts. The paper describes the methodology for approximating the p-value function, including the use of a weighted loss function and a specific weight function to emphasize the importance of small p-values. The quantile function is approximated using Monte Carlo simulation, which allows for the estimation of the true p-values. The results show that the approximations are highly accurate, with small errors even for large p-values. The paper also provides empirical illustrations using autoregressive models, demonstrating the practical utility of the approximations in real-world applications. The paper concludes that the proposed approximations are accurate and efficient, providing a practical tool for applied economists to calculate asymptotic p-values for structural-change tests. The methods are implemented in a GAUSS program, which is available for download. The paper also references related work by other authors, including Hansen (1992) and MacKinnon (1994), and discusses the broader implications of the results for nonstandard statistical distributions.This paper presents computationally convenient approximations to the asymptotic p-value functions for the Andrews and Andrews-Ploberger structural-change tests. These approximations enable accurate calculation of asymptotic p-values for these tests, which have nonstandard distributions depending on two parameters: the number of parameters tested and the range of the sample examined for the break date. The paper introduces a parametric approximation method for the p-value function, using a polynomial function combined with a leading distribution function, such as the chi-squared distribution. This approach allows for accurate and efficient computation of p-values, even for small p-values, and is applicable in a wide range of nonstandard statistical contexts. The paper describes the methodology for approximating the p-value function, including the use of a weighted loss function and a specific weight function to emphasize the importance of small p-values. The quantile function is approximated using Monte Carlo simulation, which allows for the estimation of the true p-values. The results show that the approximations are highly accurate, with small errors even for large p-values. The paper also provides empirical illustrations using autoregressive models, demonstrating the practical utility of the approximations in real-world applications. The paper concludes that the proposed approximations are accurate and efficient, providing a practical tool for applied economists to calculate asymptotic p-values for structural-change tests. The methods are implemented in a GAUSS program, which is available for download. The paper also references related work by other authors, including Hansen (1992) and MacKinnon (1994), and discusses the broader implications of the results for nonstandard statistical distributions.