This paper examines the abnormal accruals of a sample of 94 firms that reported debt covenant violations in annual reports. We expect debt covenant restrictions to influence accounting choices in the year preceding and the year of violation. Time-series and cross-sectional models are used to estimate ‘normal’ accruals. In the year prior to violation, both models indicate that ‘abnormal’ total and working capital accruals are significantly positive. In the year of violation, there is evidence of positive abnormal working capital accruals after controlling for management changes and auditor going concern qualifications.