A new method for estimating the power-law constitutive parameters from experimental data is presented. The algorithm is well suited to real time computation because the integrals employed can be continuously updated with new data. The method requires less computation than least squares fitting and avoids the problem of excessive weight being put on low amplitude data that is present in logarithmic least squares fitting. Because the method employs integrals, it smooths noise in the data. The method can also be extended to linear plus power-law fitting.