This study reports on the development and evaluation of a protocol for testing DEXA systems, which can be incorporated into a routine medical physics/engineering service. Methodologies are reported for (1) scatter measurement, (2) estimation of reference dose and (3) enquiry into potential sources of overexposure. Results show that fan-beam and cone-beam systems require shielding if the walls or operator are within 1 m of the table. Patient reference dose was estimated using the dose-area product (DAP). This varied over an order of magnitude from 2 to 36 mGy cm2 in the range of systems studied. The inquiry into potential sources of overexposure revealed some weaknesses in current practise including a design which prevents the inclusion of DAP metres; beam non-uniformity; incorrect specification of patient dose and the risk of overexposure when tube operates during patient positioning.