The pre-treatment process for the determination of ultra trace elements such as heavy metals is one of major factors affecting the reproducibility as well as the uncertainty of the data. Although the ICP-MS instrument has been reported as a reasonable analytical tool, it has also not been providing clear solutions on these difficulties in the analysis of cigarette smoke. ICP-MS equipped with a collision/reaction cell (CRC) or octopole reaction system (ORS) makes the reduction of particularly argon-containing polyatomic interferences possible by several orders of magnitude. The CRC is pressurised with a gas or a mixture of gases to reduce or eliminate the interfering polyatomic species. In side cell occurs collisional dissociation and gas phase chemical reaction. For the CRC, an He-H2 mixture was used for reduction of argon-based polyatomic interferences on chromium, arsenic and selenium. In this study using ORS, it was studied how to remove spectral interference and concentrate mainstream smoke in solvent and how to increase reproducibility applying method of direct analysis. In this study, we compared three different pre-treatment methods, block digestion, microwave digestion and solvent injection, coupled with ORS-ICP-MS for the quantification of heavy metals in mainstream smoke obtained from 3R4F reference cigarettes, and evaluated those efficiencies in the recovery, repeatability and reproducibility. In all methods, volatile heavy metals, As, showed the highest CV value, and Cd showed the lowest one. However, the solvent injection among the three methods showed the most stable recovery at 96% and a CV value of the reproducibility at 2.36%. This method also showed advantages in time-consumption and compatibility with ICP-MS system. The results demonstrate that the solvent injection method can be recommended as a superior pre-treatment procedure able to reduce contaminants and spectral interference as well as loss of the elements concerned.