This work focuses on the size-controlled preparation of Ni-B nanoparticles by chemical reduction of nickel acetate with sodium borohydride in the ternary reverse micelles of cetyl trimethyl ammonium bromide (CTAB)/n-hexanol/water. A response surface methodology (RSM) with four factors: CTAB/water (0.5-3.5g/g); CTAB/n-hexanol (0.2-0.8g/g); B/Ni molar ratio (1-4); and nickel salt concentration (0.075-0.525mol/L) was used to study the mean size of nanoparticles. Based on the central composite design (CCD), a total of 32 experimental tests were performed to correlate both reagent concentrations (Ni2+ and BH4-) and microemulsion compositions (surfactant/oil/water) to the size of to-be-obtained nanoparticles. The Ni-B nanoparticles were obtained with a narrow size distribution of average diameter in the range of 4.5-30.6nm. The quadratic model developed explained adequately the non-linear nature of the modelled response (R-2=0.97, precision=33.19). Also the influence of effective variables on the mean size of Ni-B nanoparticles was examined simultaneously and discussed theoretically by using 3D-surface and 2D-contour plots. Finally optimization of CCD based on desirability function was performed. The optimal conditions were found to be at the surfactant/oil/water mass ratio of 34/55/11, nickel salt concentration of 0.24mol/L, and B/Ni molar ratio of 2.23 with desirability factor of 0.98. Moreover, validation of the optimization showed that the model predictions were very close to the experimental results with slight errors (7-10%). Finally, TEM micrographs of Ni-B nanoparticles at optimum conditions showed that by applying statistical experimental design in reverse micelle technique, not only particle size but also agglomeration could be effectively controlled.