Waiting times for elective surgery are a key issue for the NHS. The principal policy response in the English NHS has been to introduce maximum waiting time targets against which performance is measured and rewarded. The aim of this paper is to identify the effect of government targets on the distribution of waiting times in the NHS. Specifically, we investigate the following questions: How does the probability of admission for any given patient vary during the time that they wait? How is the probability of admission for any given waiting time affected by the targets? Can variations in waiting times be explained by clinical, patient, or provider-level characteristics? What implications may be drawn from our results with respect to providers' managerial responses to the targets? This paper investigates these questions by applying duration analysis techniques to waiting time data from 2001/2002 and 2002/2003 for three specialties: general surgery, trauma & orthopaedics and ophthalmology. Estimation of survival functions reveals considerable variations in waiting times between specialties, operative procedures and hospitals. Hazard rates vary over time and peaks in them-high probabilities of admission-coincide with targets and change when targets change. Amongst patient characteristics, whether they are NHS or private and whether they are day or inpatient cases both influence waiting times, but other characteristics such as age, sex and ethnicity do not.