This paper examines alternative forms of match bias arising from earnings imputation. Wage equation parameters are estimated based on mixed samples of workers who do and do not report earnings, the latter group being assigned earnings of donors who share some but not all the attributes of the recipients. Regressions that include attributes not used as imputation match criteria (e.g., union status) are severely biased. Related forms of match bias arise with respect to attributes used as match criteria, but matched imperfectly. For example, an imperfect match on schooling creates bias that flattens estimated earnings profiles within low, middle, and high education groups, while creating large jumps in returns across groups. The same pattern arises in wage-age profiles. The paper provides a general analytic expression to correct match bias in regression coefficients under the assumption of conditional mean missing at random. The full sample correction approach is compared to the alternative of omitting imputed earners from the sample, with and without reweighting. Additional problems considered are bias in longitudinal analysis and the presence of dated donors.