Objective: To investigate the extent to which risk scales were used for the assessment of self-harm by emergency department clinicians and mental health staff, and to examine the association between the use of a risk scale and measures of service quality and repeat self-harm within 6 months. Design: Observational study. Setting: A stratified random sample of 32 hospitals in England. Participants: 6442 individuals presenting with self-harm to 32 hospital services during a 3-month period between 2010 and 2011. Outcomes: 21-item measure of service quality, repeat self-harm within 6 months. Results: A variety of different risk assessment tools were in use. Unvalidated locally developed proformas were the most commonly used instruments (reported in n=22 (68.8%) mental health services). Risk assessment scales were used in one-third of services, with the SAD PERSONS being the single most commonly used scale. There were no differences in service quality score between hospitals which did and did not use scales as a component of risk assessment (median service quality score (IQR): 14.5 (12.8, 16.4) vs 14.5 (11.4, 16.0), U=121.0, p=0.90), but hospitals which used scales had a lower median rate of repeat self-harm within 6 months (median repeat rate (IQR): 18.5% vs 22.7%, p=0.008, IRR (95% CI) 1.18 (1.00 to 1.37). When adjusted for differences in casemix, this association was attenuated (IRR=1.13, 95% CI (0.98 to 1.3)). Conclusions: There is little consensus over the best instruments for risk assessment following self-harm. Further research to evaluate the impact of scales following an episode of self-harm is warranted using prospective designs. Until then, it is likely that the indiscriminant use of risk scales in clinical services will continue.