Background The National Health Service (NHS) scores well internationally on access to healthcare. But access has been measured on methods likely to undersample the more disadvantaged. Social landlords have access to more disadvantaged groups and may be able to improve health outcomes for their tenants and reduce their NHS usage by simple interventions.

Methods This is a randomised controlled trial of 547 London social housing ‘general needs’ tenants over 50 years of age. Participants were given a health assessment, then split into a control group or one of two treated groups. Following early assessment 25 participants had to be withdrawn to receive intensive treatment because of currently untreated major health problems. Participants were followed up over 18 months and changes in health outcomes and NHS usage measured.

Results Compared with the control the most intensively treated group showed non-significant improvements on health outcomes but a significant reduction in NHS resource use, on conventional costings worth some £760 per person. All 25 participants transferred to the most intensively treated group after their early health assessments showed improvement on all health outcomes at final assessment, but these improvements were not statistically significant.

Conclusions Drawing a sample from disadvantaged but not the most seriously disadvantaged groups in London revealed 4.5% of the population to have very serious untreated health problems. The reason for lack of treatment was mainly non-registration with a general practitioner or psychiatric issues. Simple interventions to a targeted group were found to produce significant reductions in NHS usage and other, although non-significant, health benefits.

Trial registration number ID ISRCTN96259142.