We compared adherence to appropriateness criteria for transthoracic echocardiography in a Veterans Administration Medical Center (VAMC) and an academic practice and, within the VAMC, between physicians and mid-level providers. We reviewed 201 outpatient echocardiograms performed in the laboratory of an academic practice and 424 outpatient and inpatient studies performed at a VAMC. Echocardiographic examinations requested for indications addressed in the criteria were considered classified, and those for indications not addressed were considered unclassified. Classified studies were further rated as appropriate or inappropriate. Of 625 echocardiograms reviewed, 99 (16%) were unclassified. Approximately 80% of the indications for these could be assigned to 4 categories. Of the remaining 526 echocardiograms, indications were appropriate in 481 (91.4%) and inappropriate in 45 (8.6%). Among classified outpatient studies at the VAMC, mid-level providers requested significantly more studies for inappropriate indications than physicians (16.0% vs 7%, P = .024). There was no significant difference in the frequency of outpatient studies requested for inappropriate indications by VAMC and academic practice physicians (7.0% vs 9.5%, P = .558). The appropriateness criteria perform reasonably well at evaluating variations in use of echocardiography between health care systems and providers. The large majority of studies are requested for appropriate indications, although there is room for improvement.