To ascertain the influence of surgical delay on outcome after proximal femoral fracture in elderly females, a cohort study of patients presenting in 1987 was compared to 1989/90. Organisational changes in the intervening period were introduced to reduce delay to surgical intervention. Two hundred and eighty females aged 65 years and over presenting from the local catchment area of an acute inner-city teaching hospital were enrolled in the study. Seventy-nine patients received surgery in 1987 and 186 in 1989/90. The one year mortality was 34% and 26% respectively. The proportion receiving surgery within 24 hours rose from 34% in 1987 to 57% in 1989/90. The relative hazard of the group receiving surgery on day 2 in comparison to day 1 was 1.7 (95% CI 1.0 to 2.9) when adjusted for co-variance of age and mental score. Medically fit elderly patients presenting with proximal femoral fracture have improved survival with early surgery within 24 hours of admission. Improvements in the organisation of hospital care will result in important benefits for the increasing number of elderly females presenting with proximal femoral fracture.