The content of selenium and tellurium used in steels and alloys as alloying additives is strictly regulated by the current state standards. The methods of analysis used in state standards are often lengthy and laborious and do not provide their simultaneous determination. Development of alternative methods for the determination of selenium and tellurium in metallurgical materials using modern analytical equipment is an urgent goal. The method of atomic emission spectrometry with inductively coupled plasma (ICP-AES) is widely used in analysis of metallurgical materials. However, the components of the sample base can significantly affect the results of ICP-AES analysis. We present the results of theoretical and experimental study of the effect of the components of the base of samples of metallurgical materials (iron, nickel, chromium, molybdenum, cobalt, copper, tungsten) on the spectral analytical lines of selenium and tellurium. Thermodynamic modeling was used for theoretical prediction of the processes occurring during atomization of the analyzed solutions in argon plasma. It is shown that matrix non-spectral noise on the selenium line is attributed to chromium (at chromium concentrations above 50 mg/dm3), whereas the matrix non-spectral noise on the tellurium line is not observed. It is also shown experimentally that the spectral lines of selenium and tellurium are not free from spectral overlaps attributed to macrocomponents (iron, nickel, chromium, molybdenum, cobalt, copper and tungsten). To reduce the limits of detection and improve the reliability of ICP-AES determination of selenium and tellurium, it is necessary to separate them from the components of the sample base.