We approached the issue of surgical margins in the conservative treatment of breast cancer by examining the literature germane to four precise questions: At what distance from the macroscopic margin of the tumour should the resection margin be? To what extent do histologically clear resection margins indicate complete local control of the disease? To what extent do histologically involved margins indicate persistence of disease? and Does the local recurrence rate correlate with the status of the resection margin? We propose categorizing margin involvement into five groups (absent, focal, minimal, moderate and extensive involvement) according to strict histological criteria, and assigning increasingly aggressive subsequent treatments according to the extent of any margin involvement.