National Institute for Health and Clinical Excellence guidelines recommend the use of 'Track and Trigger' systems to identify early clinical deterioration. The Standardised Early Warning Score (SEWS) is used in the Royal Infirmary of Edinburgh. Previous work, suggested that the frequency and accuracy of SEWS documentation varied throughout the hospital. A prospective study was performed over a 14-night period looking at SEWS documentation in patients causing clinical concern requiring medical review, or triggering a SEWS of 4 (the 'trigger' score). SEWS charts were examined the following morning. In the ward arc, SEWS documentation was correct in only 21% of cases. The most frequent errors were one or more observations omitted (64%), SEWS total not calculated (55%) or incorrectly calculated (21%). Up to five errors per chart were noted. The observations most frequently omitted were respiratory rate, temperature and neurological status. In contrast, SEWS documentation was correct in 68% of patients in the combined assessment unit (CAU). This study demonstrates significant deficiencies in the overnight use of SEWS, particularly on the ward arc. This is particularly concerning as this study was limited only to patients already causing clinical concern, and highlights that basic observations are often incomplete, and the SEWS chart poorly understood and acted upon. SEWS recording and documentation was significantly better in CAU (P < 0.001, FET), where there is a dedicated, ongoing SEWS education programme for nursing and medical staff. We recommend this is rolled out across the hospital. Alternative methods of improving the use of SEWS are considered.