I developed this article at the Homeland Security Institute when I was Deputy Director/Chief Scientist. There was a strong call for improvements in U.S. Federal Government systems to detect SNM (special nuclear material), especially concerning detection at greater distances, but little consensus about what was possible versus what was realistic. Due to my experience developing such systems at the Naval Research Laboratory, I was in a position to add clarity to the discussion. This paper uses purely mathematical arguments to set limits on what types of improvements can be made, especially as concerns stand-off distance. It was published in the Journal of Homeland Security and available on their website. However, that website is currently undergoing renovation and their published articles are not readily available. I am therefore making the article available on the link below, until such time as the journal archives are again available.