Conservative breast surgery and modified radical mastectomy may, according to recent reports, yield equivalent survival. Analysis of a 1985 Canada-wide study (N = 228) compared surgeons and oncologists still recommending modified radical mastectomy (30%) with those recommending less aggressive surgery (69%) for a hypothetical stage I patient. The groups did not differ significantly in most physician characteristics, estimated survival and cure probabilities, importance of most treatment goals, uncertainty about treatment choice, or most attitudinal responses. Although equally involved with and cognizant of the value of clinical trials, the modified radical group expressed more skepticism about the ability of trial results to be transferred to practice and to take sufficient account of patient uniqueness, indicating greater focus on variation than mean results. Trial results might be more readily adopted if they are reported in accessible data-bases, incorporating patient characteristics potentially relevant to treatment choice. This would allow clinicians to individualize treatment by analyzing patient subsets of their own choosing.