Whether timing of surgery in relation to the menstrual cycle influences the survival of premenopausal women with breast cancer is disputed, We have conducted a retrospective study of 221 premenopausal women treated by one surgeon (JMM) with a median follow-up of 9.6 years (range 0.6, 22.1 years), Patients operated on 3-12 days from last menstrual period (LMP) (group 1; n = 80) were compared with patients operated on 0-2 and 13-28 days from LMP (group 2; n = 141), By univariate analysis group 1 patients were found to have superior disease-free survival (DFS; chi-squared = 4.55, p = 0.03) and overall survival (OS: chi-squared = 6.76, p = 0.009), with LMP group ranking behind nodal status (n = 136: DFS; chi-squared = 40.73, p = 0.00001, OS; chi-squared = 41.09, p < 0.00001) and tumour size (n = 216: DFS; chi-squared = 19.89, p < 0.001, OS; chi-squared = 15.56, p = 0.0001) in importance, This difference remained significant when adjusted individually for nodal status; tumour size, age and treatment groups, The difference was not found to be significant in patients with tumours less than or equal to 3 cm (n = 146: DFS; chi-squared = 0.31, p = 0.58: OS; chi-squared = 0.16, p = 0.69), but was highly significant in patients with tumours > 3 cm (n = 70: DFS; chi-squared = 9.91, p = 0.0016: OS; chi-squared = 15.74, p = 0.0001). In multivariate analysis LMP group remained significant for both DFS and OS, and was independent from nodal status, tumour size and age, These data support the need for a prospective study.