In this paper, a comparative study between three acid decomposition procedures of milk powder samples for the determination of Ca, Cu, Fe, K, and Na by optical emission spectrometry with inductively coupled plasma is presented. Utilizing a reflux system, it was possible to increase the temperature of the digester block above the boiling point of the reaction medium, avoiding the loss of analyte and the excessive evaporation of the acids during the heating. Conditions of acidity (10 mL HNO 3 /H 2 SO 4 ), temperature (350°C), and decomposition time (4 h) were defined by evaluation of residual carbon content. Acid digestion in an open system with a cold finger was compared with a conventional open system (4 h ate 120°C), as well as with microwave digestion (heating ramp to 180°C by the time 45 min), for four commercial samples of milk powder. For the three sample preparation procedures, the values of the analyte concentrations in all analyzed samples were similar. The accuracy of the acid digestion procedures for Ca, Cu, Fe, K, and Na was assessed by the analysis of a certified reference material of milk (infant formula), and good agreement between the certified values and the measured value was obtained. The limits of quantification for the acid digestion with reflux system for Ca, Cu, Fe, K, and Na, in milligram per kilogram, were 7. 48, 0.31, 1.0, 8.20, and 25.81, respectively, and for the microwave digestion were 14.0, 0.57, 1.78, 14.65, and 46.08, respectively. The decomposition procedure using a cold finger in an open system is effective, presenting the lowest residual carbon content (6 %), good accuracy, and relative standard deviation lower than 5 %, and appears to be a suitable procedure for sample preparation for routine analysis.