The statistical methods of experimental design were applied to efficiently increase the quality of sputtered niobium thin films. A large number of potential factors and interactions were explicitly taken into consideration in order to optimize the particular fabrication process. Three factors were found to influence the intrinsic stress of niobium films; five further factors affected the residual resistance ratio. The process behavior was modeled and optimized using these factors. The process description presented is more comprehensive than typically reported, and the results deviate from published data.