BACKGROUND Common sense suggests and research indicates relationships between staff factors in residential dementia care and quality of life (QOL) for residents, with poor care increasing suffering. However, we do not have a coherent picture of which staff interventions have an impact on quality of care (QOC) or resident QOL. METHODS A comprehensive search of 20 years' peer-reviewed literature using Medline, PsycINFO, Embase, PubMed, CINAHL, and the Cochrane, Campbell Collaboration identified 4,760 studies meriting full text review. Forty-six met the inclusion criteria, namely interventions in long-term facilities helping staff develop their capacity to provide better care and/or QOL for residents with dementia. Thirty-five other papers comprised an associated predictor review. RESULTS Conclusions from these limited data are further compromised because nine studies failed to measure effects on residents and only half assessed effects after the project team withdrew. Of these, excellent studies produced change over the medium (3-4 months) or longer term, including reduction in challenging behavior and restraint use but this applied only to a minority. A number of studies failed to measure effects on QOC, limiting conclusions about mechanisms underlying change. CONCLUSION In general, level of intervention required depended on the target. For outcomes like restraint use, structured education sessions with some support appear adequate. Programs to reduce pain require more support. For complicated issues like challenging behavior and increasing co-operation in showering, detailed, supportive, on-site interventions are required. Improvements in restraint and staff/resident interactions were the most promising findings. (Review registration number: PROSPERO 2014:CRD42014015224).