A technique is disclosed for classifying a population of subjects into various sub-populations for a selected biological condition. Patients are categorized in accordance with numeric scores for a affected status for the selected biological condition and a risk status for the selected biological condition. The numeric scores for an overall population are determined in advance for the selected biological condition. Medical test results, including genetic tests, and risk factors are numerically scored and may further be weighted in accordance with their relevance in determining affected status and risk. Medical test results and medical histories for individual subjects within the population may then automatically be scored in accordance with the predefined characteristics. The numerical scores for affected status and risk status may be stored in a data structure, such as a database. The numeric scores are extracted from the data structure and used to classify individuals in the population into one of a group of selected sub-populations comprising at-risk affected (ARA) and at-risk unaffected (ARU). Additional sub-populations, such as unknown risk, unaffected (URU) may also be used.