To determine whether the quality indicator 'tumour positive surgical margin following breast-conserving surgery, consistently measured the quality of breast-cancer surgery independently of the different definitions used and differences in case mix, taking statistical random variation into account. Descriptive study. Data was collected from 762 patients who underwent breast-conserving surgery for invasive or in situ carcinoma of the breast, in the period 1 July 2007 - 30 June 2008 in 1 of the 9 hospitals in the region of the Comprehensive Cancer Centre West in the Netherlands. We compared 3 definitions for 'tumour positive surgical margin': the one used by the Health Care Inspectorate, the one used by the organisation 'Zichtbare Zorg' ('transparent care') and the percentage of re-resection. For case mix correction we identified risk factors for tumour margin positivity with logistic regression. The results were presented in a funnel plot, using 95% confidence interval (CI) around the national standard of 20%. Depending on the definition used, the tumour positive surgical margin rate of the total group varied from 11 to 21%. Individual hospital rates varied by up to 19%. In situ carcinoma was associated with higher tumour positive surgical margin rates. Results differed significantly between hospitals for all 3 definitions. However, the funnel plot showed that results for most hospitals fell within the 95% CI of the standard. Whether a hospital fell within the 95% CI of the standard depended upon on the definition used and case mix correction. The lack of a single definition for the quality indicator 'tumour positive surgical margin following breast-conserving surgery' and the lack of case-mix correction undermine the validity of the indicator. Standardisation of definitions, uniform registration and the use of funnel plots can provide a more transparent insight into the quality of care.