Objective: In order to determine whether post-acute intermediate care in nursing-led inpatient units (NLUs) is effective in preparing patients for discharge from hospital we conducted a systematic review of the evidence. Review methods: The Cochrane Library, Effective Practice and Organisation of Care specialist register, Medline, Cinahl, Embase, British Nursing Index and the HMIC databases were searched for all available dates up to mid-2003. The science and social science citation indices were searched for papers that cited key works. Authors of papers were asked to identify additional research. Randomised controlled trials, controlled clinical trials, controlled before and after studies and interrupted time-series designs that compared the NLU to usual post-acute inpatient care for adults were included in the review. Studies were assessed for quality. Statistical meta-analysis on the results of controlled trials was performed. Sensitivity analyses were conducted to determine the impact of methodological quality on conclusions. Outcomes: Outcomes considered were mortality, institutionalisation after discharge, functional status early readmission, length of inpatient stay and cost. Results: Nine random or quasi-random controlled trials involving 1669 patients were reviewed. Quality was variable. The mean age of patients in all studies was over 70 years. There was no statistically significant difference in inpatient mortality between NLU and usual inpatient care (OR 1.10, 95% CI 0.56–2.16). The NLU was associated with reduced odds of discharge to institutional care (OR 0.44 95% CI 0.22–0.89), better functional status at discharge (SMD 0.37, 95% CI 0.20–0.54) and reduced odds of early readmission (OR 0.52 95% CI 0.34–0.80). Length of stay until discharge home was increased by 5.13 days (WMD) (95% CI–0.5–10.76 days). At longest follow up (3–6 months) there was no statistically significant difference in the proportion of patients in institutional care (OR 0.97, 95% CI 0.60–1.58). The results were not generally sensitive to study quality. Conclusions: The NLU successfully functions as a form of intermediate care, so far there is no evidence of adverse outcome from the lower level of routine medical care. However, more research is required to confirm safety. Patients discharged from NLUs have higher levels of function although it is unclear if the benefit is simply a product of an increased stay. There is no evidence of benefit over the longer term.