Summary: An important assumption used in the analysis of line transect data for animal abundance estimation is that the perpendicular sighting distances of detected animals to transect lines are measured accurately. However, in many line transect surveys, the sighting distances are likely to be subject to measurement errors. We investigate the effect of measurement errors on animal abundance estimation and show that measurement errors cause systematic bias, which cannot be reduced by simply increasing the survey effort. Estimators that correct for measurement errors using the method of moments by assuming an exponential power series detection function and knowledge of only a few moments of the error distribution are proposed.