A process-oriented quality care audit was performed in a large metropolitan hospital emergency radiology facility with an annual volume of over 50,000 examinations. One aspect of the audit dealt with errors found among interpretations by radiology residents, the initial interpreters of x-ray studies. Misinterpretations were identified by staff radiologists, who checked all examinations and countersigned the reports. Error rates were correlated with duration of training and were separated as to significance and whether the errors were false-negative (omission) or false-positive (commission). The false-positive to false-negative ratio was 27:73% which is in agreement with previous studies. For all cases of errors, the significance of change in interpretation was high in 20%, moderate in 29% and low in 51%. The effect of inadequate clinical history on the rate and significance of interpretation errors was also determined. When clinical information was inadequate, the significance was high in 27%, moderate in 40% and low in 33%.