In this thesis, a statistical analysis of the data obtained from the design of experiment was carried out. A two-factor experiment plan was applied where processing parameters, i.e. input factors, were changed at two levels with three repetitions of the combination of factor levels. Processing parameters, i.e. input factors changed at two levels are: spindle rotation speed (315 rpm and 630 rpm) and feed (0,25 mm/r to 0,45 mm/r), with a constant cutting depth of 1 mm. For the experiment, samples of general structural steel were used, which were processed on the Prvomajska TNP 180 universal lathe. The aim of the experiment was to see if the spindle speed and feed affect the surface roughness and to obtain regression models for predicting the surface roughness. Computer program MS Excel is used for the statistical analysis of the experimental data. After the statistical analysis, it can be concluded that the feed has a significant influence on the surface roughness, while the spindle speed and the interaction of the spindle speed and the feed do not significantly affect the surface roughness. Regression models were also obtained, with a high coefficient of determination (R2), for predicting the surface roughness.