Background Preventing cognitive impairment in older adults is a public health priority. Although multidomain interventions have shown promise as preventive strategies, the optimal combination of interventions remains unclear. This network meta-analysis aimed to compare and rank the relative efficacy of single-domain and multidomain lifestyle interventions for the prevention of cognitive impairment in older adults who are cognitively unimpaired. Methods We did a systematic review and network meta-analysis of randomised controlled trials (RCTs) published in PubMed and Embase from inception until the date of our search on May 7, 2024 following a preregistered protocol in PROSPERO (CRD42024601975). We included RCTs in older adults who are cognitively unimpaired evaluating lifestyle interventions targeting diet, physical exercise, cognitive training, social activity, and health education, either alone or in combination. The primary outcome was global cognition, analysed using random-effects network meta-analysis, reporting standardised mean differences (SMDs) and 95% CIs, and compared against health education, active control, or no intervention. Subgroup analyses explored potential age-related differences and the effect of intervention duration. Risk of bias was assessed using Cochrane Risk of Bias 2, and publication bias was evaluated by assessing funnel plot asymmetry. Findings Of the 10 200 citations identified and 1183 full texts screened for eligibility, we identified 109 eligible RCTs, including 23 010 participants (median age 70⋅1 years [IQR 68⋅7-73⋅8], 14 957 [65%] female and 8053 [35%] male). Compared with health education, significant improvements in global cognition were found for physical exercise and cognitive training combined (SMD 0⋅26 [95% CI 0⋅10-0⋅42; p=0⋅0011); cognitive training alone (SMD 0⋅21 [0⋅08-0⋅33]; p=0⋅00092); diet, physical exercise, cognitive training, and health education combined (SMD 0⋅14 [0⋅02-0⋅27]; p=0⋅028); and physical exercise alone (SMD 0⋅14 [0⋅05-0⋅22]; p=0⋅0014). Random-effects models using active control and no intervention as comparators yielded similarly significant effects for the aforementioned interventions, with effect sizes in the same order. Risk of bias was high in 44 (40%) studies, and publication bias was suggested in studies comparing interventions with health education. Interpretation Several single-domain and multidomain lifestyle interventions are efficacious at modulating global cognition in older adults who are cognitively unimpaired, with the combination of physical exercise and cognitive training demonstrating the strongest effect. Combining lifestyle interventions might enhance efficacy, but increased number of domains does not automatically translate into greater cognitive benefits. These findings support lifestyle interventions as key components of prevention strategies; however, their optimal combination requires further investigation.