The recovery of a recombinant intracellular protein, cytochrome b5, from Escherichia coli TB1 cells was carried out by bead mill disintegration in a discontinuous small-scale instrument. This process was optimized by the use of experimental factorial design. Several parameters were studied: operating time, amount and size of beads, cellular suspension concentration, and presence of toluene and lysozyme. For the experimental conditions used, only the time of treatment and bead load had significant effects. The optimal values of these variables were found by applying the response surface methodology.