Weight loss, malnutrition and dehydration are common problems for people with dementia. Environmental modifications such as, change of routine, context or ambience at mealtimes, or behavioural modifications, such as education or training of people with dementia or caregivers, may be considered to try to improve food and fluid intake and nutritional status of people with dementia. Primary: To assess the effects of environmental or behavioural modifications on food and fluid intake and nutritional status in people with dementia. Secondary: To assess the effects of environmental or behavioural modifications in connection with nutrition on mealtime behaviour, cognitive and functional outcomes and quality of life, in specific settings (i.e. home care, residential care and nursing home care) for different stages of dementia. To assess the adverse consequences or effects of the included interventions. We searched the Specialized Register of Cochrane Dementia and Cognitive Improvement (ALOIS), MEDLINE, Eembase, PsycINFO, CINAHL, ClinicalTrials.gov and the World Health Organization (WHO) portal/ICTRP on 17 January 2018. We scanned reference lists of other reviews and of included articles. We included randomised controlled trials (RCTs) investigating interventions designed to modify the mealtime environment of people with dementia, to modify the mealtime behaviour of people with dementia or their caregivers, or both, with the intention of improving food and fluid intake. We included people with any common dementia subtype. Two review authors independently selected studies, extracted data and assessed the risk of bias of included trials. We assessed the quality of evidence for each outcome using the GRADE approach. We included nine studies, investigating 1502 people. Three studies explicitly investigated participants with Alzheimer's disease; six did not specify the type of dementia. Five studies provided clear measures to identify the severity of dementia at baseline, and overall very mild to severe stages were covered. The interventions and outcome measures were diverse. The overall quality of evidence was mainly low to very low.One study implemented environmental as well as behavioural modifications by providing additional food items between meals and personal encouragement to consume them. The control group received no intervention. Differences between groups were very small and the quality of the evidence from this study was very low, so we are very uncertain of any effect of this intervention.The remaining eight studies implemented behavioural modifications.Three studies provided nutritional education and nutrition promotion programmes. Control groups did not receive these programmes. After 12 months, the intervention group showed slightly higher protein intake per day (mean difference (MD) 0.11 g/kg, 95% confidence interval (CI) -0.01 to 0.23; n = 78, 1 study; low-quality evidence), but there was no clear evidence of a difference in nutritional status assessed with body mass index (BMI) (MD -0.26 kg/m² favouring control, 95% CI -0.70 to 0.19; n = 734, 2 studies; moderate-quality evidence), body weight (MD -1.60 kg favouring control, 95% CI -3.47 to 0.27; n = 656, 1 study; moderate-quality evidence), or score on Mini Nutritional Assessment (MNA) (MD -0.10 favouring control, 95% CI -0.67 to 0.47; n = 656, 1 study; low-quality evidence). After six months, the intervention group in one study had slightly lower BMI (MD -1.79 kg/m² favouring control, 95% CI -1.28 to -2.30; n = 52, 1 study; moderate-quality evidence) and body weight (MD -8.11 kg favouring control, 95% CI -2.06 to -12.56; n = 52, 1 study; moderate-quality evidence). This type of intervention may have a small positive effect on food intake, but little or no effect, or a negative effect, on nutritional status.Two studies compared self-feeding skills training programmes. In one study, the control group received no training and in the other study the control group received a different self-feeding skills training programme. For both comparisons the quality of the evidence was very low and we are very uncertain whether these interventions have any effect.One study investigated general training of nurses to impart knowledge on how to feed people with dementia and improve attitudes towards people with dementia. Again, the quality of the evidence was very low so that we cannot be certain of any effect.Two studies investigated vocal or tactile positive feedback provided by caregivers while feeding participants. After three weeks, the intervention group showed an increase in calories consumed per meal (MD 200 kcal, 95% CI 119.81 to 280.19; n = 42, 1 study; low-quality evidence) and protein consumed per meal (MD 15g, 95% CI 7.74 to 22.26; n = 42, 1 study; low-quality evidence). This intervention may increase the intake of food and liquids slightly; nutritional status was not assessed. Due to the quantity and quality of the evidence currently available, we cannot identify any specific environmental or behavioural modifications for improving food and fluid intake in people with dementia.