Evidence-based guidelines for the diagnosis of venous thromboembolism (VTE) have been recommended, yet the adoption of such guidelines into daily practice is unknown. The purpose of this study was to describe the current practices in the diagnosis of VTE. Medical records of 1161 adult patients who underwent lower extremity venous duplex scans (VDS), chest computerized tomographic (CT) angiography, or ventilation and perfusion (V/Q) scans during a 6-month period were retrospectively reviewed in an academic medical center. Patients who were first diagnosed by CT or V/Q scan still underwent a VDS. Nine patients at high risk had incomplete CT scans, yet no further tests were performed. Five pregnant patients had CT scans as the initial test instead of being screened with VDS or V/Q scanning. Inappropriate use of imaging tests was documented. The recommended guidelines of using clinical probability and D-dimer as the initial screening tests for VTE diagnosis were underused.