We propose a new estimation methodology to address the presence of covariate measurement error by exploiting the availability of spatial data. The approach uses neighboring observations as repeated measurements, after suitably controlling for the random distance between the observations in a way that allows the use of operator diagonalization methods to establish identiﬁcation. The method is applicable to general nonlinear models with potentially nonclassical errors and does not rely on a priori distributional assumptions regarding any of the variables. The method’s implementation combines a sieve semiparametric maximum likelihood with a ﬁrst-step kernel conditional density estimator and simulation methods. The method’s eﬀectiveness is illustrated through both controlled simulations and an application to the assessment of the eﬀect of pre-colonial political structure on current economic development in Africa.