Abstract
Favourable convergence properties are demonstrated for a simple iterative method of nonlinear least-squares estimation. The method is applicable when the non-linear model is linear with respect to each of the parameters as soon as the other ones are constants.