The speeds of screen -film combinations available in Canada have been measured at 75 kVpwith appropriately filtered beams. The influence on speeds of kVp, attenuation of thefront of the cassette, age and use of screens, reciprocity failure and film variations frombatch to batch is surveyed. The significance of the observed variations in speed (inselecting screen -film combinations) is considered. A set of radiographs of head and bodyphantoms has been used to examine the criteria which radiologists use in defining a 'good'radiograph and from this a qualitative assessment has been made of the priorities inselection of screen -film combinations.IntroductionThe selection of screen -film combinations involves a choice between several conflictingfactors, such as eye appeal of the radiograph, accuracy of diagnosis, dose to the patient,and efficiency and economy of operation of a radiology department. In helping radiologistsand technologists to make a selection we have attempted to answer their questions about thephysical properties of screens and films. Examples of our measurements and results arereported here and their significance to the radiologist is considered. In return radiol-ogists have shown us from sets of test radiographs of phantoms what they consider to be'good' radiographs and have tried to define the parameters they consider important in imagequality. From this we have obtained a qualitative assessment of the relative prioritiesgiven to these parameters in selecting a screen -film combination for general radiography.MethodsSpeeds of screen -film combinations (defined here as the reciprocal of the exposure inmR needed to produce a density of 1.0) were measured using a calibrated three -phase 12pulse generator. Water, lucite and TEMEX water -equivalent rubber phantoms (J. Girdler andSons Ltd., London, U.K.) were used to simulate clinical conditions and provide a suitablyfiltered beam. A simple jig was used to ensure consistent collimation of the X -ray beam.The X -ray exposure of the cassette was measured directly above the front face of thecassette using a pancake -type calibrated ionisation chamber dosimeter. The speeds measuredusing this arrangement were the same as those obtained when the cassette was used with amoving grid in a Bucky tray.Films were processed in DuPont XMD chemistry in a Kodak M6AN processor at 92` F.