We consider nonparametric estimation of a regression curve when the data are observed with multiplicative distortion which depends on an observed confounding variable. We suggest several estimators, ranging from a relatively simple one that relies on restrictive assumptions usually made in the literature, to a sophisticated piecewise approach that involves reconstructing a smooth curve from an estimator of a constant multiple of its absolute value, and which can be applied in much more general scenarios. We show that, although our nonparametric estimators are constructed from predictors of the unobserved undistorted data, they have the same first order asymptotic properties as the standard estimators that could be computed if the undistorted data were available. We illustrate the good numerical performance of our methods on both simulated and real datasets.1. Introduction. We consider nonparametric estimation of a regression curve m(x) = E(Y |X = x) when X and Y are observed with multiplicative distortion induced by an observed confounder U . Specifically, we observeX,Ỹ and U , whereỸ = ψ(U ) Y ,X = ϕ(U ) X, ψ and ϕ are unknown functions and U is independent of X and Y . This model is known as a covariate-adjusted regression model. It was introduced by Şentürk and Müller (2005a) to generalize an approach commonly employed in medical studies, where the effect of a confounder U , for example body mass index, is often removed by dividing by U . Motivated by the fibrinogen data on haemodialysis patients, whereỸ was fibrogen level,X was serum transferrin level, and U was body mass index, Şentürk and Müller (2005a) pointed that although it is often reasonable to assume that the effect of U is multiplicative, it does not need to be proportional to U , and a more flexible model is obtained by allowing for distortions represented by the functions ϕ and ψ. More generally, this model is useful to describe the relationship between variables that are influenced by a confounding variable, and see if this relationship still exists once the effect of the confounder has been removed.A number of authors have suggested estimators of the curve m in various