In this paper we study the impact of misreported treatment status on the estimation ofcausal treatment effects. We characterise the bias introduced by misclassification on theaverage treatment effect on the treated under the assumption of selection on observables.Although the bias of matching-type estimators computed from misclassified data cannot ingeneral be signed, we show that the bias is most likely to be downward if misclassificationdoes not depend on variables entering the selection-on-observables assumption, or onlydepends on such variables via the propensity score index. We extend the framework tomultiple treatments. We provide results to bound the returns to a number of educationalqualifications in the UK semi-parametrically, and by using the unique nature of our datawe assess the plausibility for the two biases from measurement error and from omittedvariables to cancel out.