Background: A nutrition risk screening (NRS) tool (after Ward et al., 1998) was introduced into an older adult mental health service in England, in 1999. It was considered to be the most practical of the tools available at the time. Trust standards for its use are that inpatients should have NRS completed within three days of admission; and should be weighed on admission and then weekly during their stay. The aims of these audits were to assess compliance with trust standards, and the effect of inpatient care on patients’ weight.Methods: Audits included patients discharged to a residential address, in 6 months sampled across the year, to account for any seasonal effects. Notes were audited for the recording of NRS and weights, length of stay and patient demographics. Paired two‐tailed t‐test measured the significance of the difference between admission and discharge weights.Results: Most results were consistent over three audits (Table 1) and standards for NRS and weight monitoring were met. The higher proportions of women were expected in this age group. Over 30% of patients were at increased nutritional risk on admission. There were twice as many thin people as obese people. Inpatient care resulted in statistically significant weight increase. Length of stay was ten times the NHS average (Black & Pearson, 2002). Results of audits conducted in 2002, 2004 and 2007 (n (%)). 2002 2004 2007 Months audited Jan Feb May Jun Sept Oct Mar Apr Jul Aug Nov Dec Jan Mar May Jul Sept Nov Discharges audited 178 (84) 128 (85) 127 (83) Gender (%) women/men 64/36 66/34 66/34 Mean (SD) age (years) 78.7 (8.3) 77.3 (7.3) 78.5 (7.9) Patients screened for nutritional risk 96 95 96 Nutritional risk scores: low 39 (22) 18 (10) 60 (47) medium 36 (28) 22 (17) 77 (61) high 114 (64) 36 (28) 9 (7) Weight monitored (%) 98 98 100 Body mass index (kg/m2) <20 87 (49) 41 (23) 15 (8) 20 – 25 20 (16) 64 (50) 32 (25) 26 – 30 10 (8) 20 (16) 65 (51) >30 31 (17) 33 (26) 9 (7) Mean (SD) admission weight (kg) 65.2 (16.0) 63.0 (14.7) P = 0.000 Mean (SD) discharge weight (kg) 64.0 (14.0) 62.0 (15.9) P = 0.001 Paired 2 tailed t‐test 63.6 (16.7) 63.9 (15.4) P = 0.000 Mean (SD) length of stay (days) 65 (51) 62 (50) 88 (146) Discussion: NRS has been repeatedly recommended for at least a decade, most recently by NICE (2006). Much effort is put into validating NRS tools but little into ensuring their use. This audit is among the first reports of long term, routine use of NRS. More work is needed to show that the statistically significant weight gain found here has clinical, functional or economic significance or importance. Changing nutritional status is a relatively long‐term exercise so the influence of the length of stay needs further exploration.Conclusions: NRS and weight monitoring can be incorporated into routine clinical care. This is the first report of their long‐term use in older people's psychiatry and, perhaps, in any UK health setting. The implications of the findings merit more detailed analysis.References Black, D. & Pearson, M. (2002) Average length of stay, delayed discharge and hospital congestion. Br. Med. J. 325, 610–611.NICE. (2006) Nutritional support in adults: oral nutritional support, enteral tube feeding and parenteral nutrition. Available at http://www.nice.org.uk/cg032 (Accessed on 25 March 2006).Ward, J. et al. (1998). Development of a screening tool for assessing risk of undernutrition in the community. J. Hum. Nutr. Diet.11, 323–330.