It is one matter to decide to "go joint" - and another to successfully bring it about. Although the Meteorology Project, which was developing numerical weather prediction models at the Institute for Advanced Study in Princeton, had always been a jointly funded and supported project of the Weather Bureau, Navy, and Air Force, taking the models into the operational realm would require much more inter-weather service cooperation than had been required in the past. Further, there were many non-modeling challenges that needed to be faced: securing a computer to run the models, finding and outfitting an appropriate physical space, dealing with funding and manpower issues, and melding three different institutional cultures. All of these concerns needed to be addressed with minimal in-fighting and a setting aside of personal agendas. Furthermore, because NWP was such a new endeavor, participants were forced to deal with a federal bureaucracy that did not always grasp the complex needs of this very special undertaking. While successfully running models were obviously crucial to operational NWP - all the supporting issues were critical as well. This paper will discuss the other problems that faced the nascent Joint Numerical Weather Prediction Unit and how members of all the participating services worked to solve them.