This work reports, for the first time, a detailed investigation on the effect of experimental parameters that affect the extraction induced by emulsion breaking. In this study, we evaluated the influence of the type (Triton X-100, Triton X-114, Tween 20 and sodium dodecylsulfate) and concentration of the surfactant employed as emulsifying agent, as well as the type (HCl or HNO3) and concentration of the mineral acid employed for the extraction of As, Cr, Cu, Mn, Pb and V of spiked mineral oil. The metals were added to the mineral oil in the form of organometallic compounds. In all cases, the emulsion breakdown was induced by heating at 90 degrees C in a thermostatic bath and the extracts were analyzed by graphite furnace atomic absorption spectrometry (GF AAS). The time required to achieve emulsion breakdown and the extraction efficiency were taken as responses for evaluation. The obtained results showed that the extraction of metals was more efficient when Triton X-100 or Triton X-114 were employed in the preparation of the emulsions. Both HCl and HNO3 provided quantitative extraction of the analytes when combined with these surfactants, except in the case of As, in which the extraction efficiencies were in the range of 22-60%. However, the emulsions prepared with HCl needed longer times to be broken down. Samples of mineral oils (new and used) from vacuum pumps were analyzed and the results compared with those obtained with the model sample. Only As and V could not be determined in the real samples.