Unexpectedly we have found large numbers of chronically ill Borrelia burgdorferi PCR- and seropositive patients in Houston, Texas, a zoonotically 'non-endemic' area. In order to understand this finding prior to sufficient data availability, we chose to examine critically currently accepted but troublesome 'Lyme disease' concepts. Our method was to analyze each foundation 'Lyme disease' premise within the context of available medical and veterinary literature, then to reconstruct the disease model consistent with the preponderance of that data. We find the present conceptualization of the illness seriously truncated, with a high likelihood of two distinct but connected forms of human B. burgdorferi infection. The yet-unrecognized form appears to have a broader clinical presentation, wider geographic distribution, and vastly greater prevalence. We conclude that 'Lyme disease' currently acknowledges only its zoonosis arm and is a limited conceptualization of a far more pervasive and unrecognized infection state that must be considered a global epidemic.