Statistical methods have been used to design an efficient experimental study of the tellurium electrowinning process where current density, electrolyte flow rate, temperature, free NaOH and initial Te concentration were varied simultaneously. Regression analysis was used to quantify the relative impact of each process operating parameter on the purity of electrowon Te. Empirical models show that significant interactions between process parameters allow for the minimization of Se and Pb co-deposition. Behaviour predicted by the models has been substantiated with voltametric and amperochronometric observations, confirming that statistical design of experiments effectively mapped the process responses.