Background: The NSF specifies that mental health service providers should have a locally agreed proforma for assessing risk. Risk assessment proformas (RAPs) currently in use vary considerably in both structure and content. This study describes some similarities and differences.Method: We requested a copy of each provider trust's RAP. These were assessed across a range of structural characteristics, including layout, design, data coding, and were also studied for their content.Results: Some consistency was seen in the themes addressed: suicide/self harm, risk to others (notably violence) and vulnerability were well represented. Several additional themes were notable by their absence. There was general conformity in the way in which data was coded, with most using tick box categories. However, there were striking differences in layout quality and usability and many common problems were identified.Conclusions: RAPs as currently used in NHS mental-health practice vary in structure, content, length and quality. We question (i) the most common approaches in recording risk data in RAPs, and (ii) whether the focus on three domains of risk acts to neglect other, less dramatic, areas. Risk assessment procedures need to be critically evaluated and the introduction of standardized tools would be advantageous in this respect.Declaration of interest: This work was supported by NHS R&D Support funding awarded to Hertfordshire Partnership NHS Trust.