Performance of Automatic Exposure control in several digital mammography systems (both CR and DR) were evaluated. The test procedure applied was that proposed by the European Guidelines, which requires measurement of contrast-to-noise ratio (CNR) produced by 0.2 mm Al superimposed on variable polymethyl-methacrylate (PMMA) thicknesses. PMMA layers were exposed by full automatic techniques and average glandular dose (AGD) determined. Results demonstrated that AGD values keep below the acceptable limits for all CR and DR systems at almost all equivalent breast thicknesses. CNR absolute values per each thickness can significantly vary, depending on characteristics of each type of equipment. The application of limits suggested by the European Guidelines for CNR variation with reference to 50 mm PMMA causes systematic failure for equivalent breast thickness above 50 mm; this may be due to the contradiction between the principles followed by the manufacturers to designed the AEC (signal constancy) and those proposed by the European protocol (CNR constancy).