The use of active personal dosemeters (APD) in interventional radiology was evaluated by Working Group 9 (Radiation protection dosimetry of medical staff) of the CONRAD project, which is a Coordination Action supported by the European Commission within its sixth Framework Programme. Interventional radiology procedures can be very complex and they can lead to relatively high doses to personnel who stand close to the primary radiation field and are mostly exposed to radiation scattered by the patient. For the adequate dosimetry of the scattered photons, APDs must be able to respond to low-energy [10-100 keV] and pulsed radiation with relatively high instantaneous dose rates. An intercomparison of five APD models deemed suitable for application in interventional radiology was organised in March 2007. The intercomparison used pulsed and continuous radiation beams, at CEA-LIST (Saclay, France) and IRSN (Fontenay-aux-Roses, France), respectively. A specific configuration, close to the clinical practice, was considered. The reference dose, in terms of Hp(10), was derived from air kerma measurements and from the measured and calculated energy distributions of the scattered radiation field. Additional Monte Carlo calculations were performed to investigate the energy spectra for different experimental conditions of the intercomparison. The results of this intercomparison are presented in this work and indicate which APDs are able to provide a correct response when used in the specific low-energy spectra and dose rates of pulsed X-rays encountered in interventional radiology.