Objective: to develop consensus guidelines for management of delirium and to assess their effectiveness in improving the outcomes and process of care in delirium. Method: guidelines for delirium were developed following a literature search and a formal multi-disciplinary consensus process using a two-stage Delphi technique. The process and outcomes of patients with delirium were then observed in a ‘before’ (211 patients) and an ‘after’ study (125 patients). Three levels of intervention were made in the ‘after’ study. (i) Feedback of baseline data only (low intensity intervention). (ii) As in (i), but also formal distribution of the guidelines to nurses and doctors (medium intensity intervention). (iii) As in (ii), but in addition the guidelines were reinforced with teaching sessions for the nurses and doctors (high intensity intervention). Setting: older people (aged over 65 years) with delirium admitted to acute medical or acute elderly care wards in five district general hospitals in England. Results: only in the high intervention group was there an improvement in process and outcome of care, but this failed to reach statistical significance. Conclusion: delirium is a poorly managed condition in older people and guidelines alone fail to improve the process and outcomes of care.