In this article, it is argued that current shielding methodologies can lead to an over-shielding of low-energy X-ray imaging facilities. Measurements of scattered air kerma rates in a mammography installation, including the effect of an anthropomorphic phantom of the patient, are presented. Detailed Monte Carlo calculations that support the measured results are also shown. Calculations for a mobile mammography installation, including the 'patient transmission factor', are presented and results are discussed.