The growth of violet light emitting diodes (LEDs) was optimized using a statistical design of experiment (DOE) approach and several important interaction effects were found. The DOEs studied the effect of several variables on the well layer, the barrier layer, and the pAlGaN cladding layer. These variables included the gallium flow rate, the indium flow rate, the growth temperature, and the growth time for the well layer, the ammonia flow in the active region, the barrier growth time, and the Si doping of the barrier, as well as the growth time, growth temperature and Mg doping of the pAlGaN cladding layer. The LEDs were optimized based on combinations of several responses from photoluminescence and electroluminescence measurements. An overall process desirability was obtained, based on achieving the desired wavelength and maximizing the PL intensity and optical output power. Significant interactions between variables played a major role in the optimization of optical output power as well the emission wavelength. The understanding of these interactions led to the optimization of the LEDs both by improvements in the structure and improvements in the quality of the layers. Several of the interactions will be explained based on kinetic models of GaN growth by MOCVD.