Empirical and survey evidence suggest that firms often manipulate reported numbers to avoid debt covenant violations. The theoretical literature, by and large, has ignored the consequences of this phenomenon on debt contracting. Departing from a standard debt financing setting with continuation decisions based on reported contractible signals, we study how firms ability to manipulate reports affect the design of debt contracts. The model generates an array of novel (and perhaps surprising) empirical predictions regarding the optimal covenant, the interest rate, the efficiency of the continuation/liquidation decisions, and the likelihood of covenant violations.