Objective: To monitor key processes and outcomes in joint health and social services community psychogeriatric teams.Design: Six month follow-up of new referrals to 4 teams in Cambridge. Data collected from structured clinical assessment forms at baseline and interviews with keyworkers at follow-up. Data for groups with and without dementia were compared.Setting: Two rural and two urban teams in area with relatively large long stay inpatient facility but low independent sector provision. All were routine service teams, rather than run for research purposes.Main outcome measures: The main outcome measures were survival, institutionalisation, key worker assessments of avoidable admissions, appropriateness of placement, unmet needs, carer stress and global outcome for patient and carers.Results: Rates of referral to urban teams were double rural rates. Around forty percent of the dementia group had a social worker as assessor and keyworker. The dementia group was significantly more dependent and received more informal and formal care. After six months, only 54% of the dementia group were alive and living outside institutional care, compared to 79% in the functionally ill group. Unplanned admissions to hospitals or homes were rarely judged avoidable by keyworkers. Outcomes for carers were judged poor in 15% of both groups, and 13% and 11% of carers respectively were judged to be under severe stress. Unmet needs were more common in the dementia group, and related principally to residential care and carer respite.A number of measures, including evidence of geographical inequity and identification of unmet needs, provided an important contribution to local policy development. Further work is needed on the validity of keyworker assessments of carer stress, given the findings of the carer interview sub-study.