Vehicle voice recognition systems have become an essential tool for hands free communication. As such, it has become more and more important to have reliable, consistent voice recognition in a vehicle. Vehicle voice recognition system performance is based on a variety of factors, including the speakers' gender & background noise. Male and female voice characteristics are inherently different, and some of these variations are investigated in this work. In this work, three vehicles have been tested during five different steady state road conditions (70 mph, 45 mph, Idle HVAC off, Idle HVAC on, vehicle off). Twelve speakers (six male and six female) were recorded announcing twenty mono- and multi-syllable call commands. Each speaker was recorded three times for repeatability, along with the vehicle voice recognition system response. Based on the resulting success rates, the least-recognized commands were synthesized to resemble the best detected commands from the different genders. These customized commands were then presented to the vehicle's voice recognition system in a controlled environment. Results show promise in the removal of gender based voice recognition variability and how the background noise can be dealt with by having a good noise cancellation algorithm.