RATIONALE: Early childhood presents an important opportunity for obesity prevention, an international public health priority. Interventions aiming to prevent obesity typically promote healthy diets or activity levels (physical activity, sedentary behaviour), or both. There is uncertainty over which approaches are more effective. This is one of a suite of three reviews addressing interventions for preventing obesity in children, each focusing on different age groups up to 18 years. These reviews replace and update a 2019 Cochrane review on interventions for preventing obesity in children from birth to 18 years. OBJECTIVES: To assess the effects of interventions that aimed to prevent obesity in children aged two to four years by changing dietary intake or activity levels, or both, on body mass index (BMI), BMI z-score (zBMI), BMI percentile, and serious adverse events. SEARCH METHODS: We searched CENTRAL, MEDLINE, Embase, six other databases, and two trial registers, together with reference checking, citation searching, and contact with study authors to identify eligible studies. The latest search date was 7 February 2023. ELIGIBILITY CRITERIA: We included randomised controlled trials (RCTs) in children aged two to four years, comparing dietary or activity interventions (or both combined) to prevent overweight or obesity versus no intervention, usual care, or another eligible intervention, in any setting. Studies had to measure outcomes at a minimum of 12 weeks post-baseline. There were no language restrictions. OUTCOMES: Our outcomes were BMI, zBMI, BMI percentile, and serious adverse events. RISK OF BIAS: We used version 2 of the Cochrane risk of bias tool (RoB 2) to assess included RCTs. SYNTHESIS METHODS: Working independently, two authors screened studies, extracted data, and conducted risk of bias and GRADE assessments. We conducted random-effects meta-analyses stratified by type of intervention and follow-up duration. INCLUDED STUDIES: We included 67 studies (36,601 participants), with 56 studies (21,404 participants) pooled in the meta-analyses. Sixty-three studies were conducted in high-income countries. Study settings were split between childcare (26 studies; 39%), home (23 studies; 34%), and home plus childcare settings (10 studies; 15%). Few studies were conducted in the community (three studies; 5%) or healthcare settings (five studies; 7%). Most studies compared a combined dietary/activity intervention with a control group. SYNTHESIS OF RESULTS: Dietary interventions versus control Dietary interventions may have little to no effect on BMI at short-term follow-up (mean difference (MD) 0.00, 95% confidence interval (CI) -0.28 to 0.28; 1 study, 94 participants; low-certainty evidence). At medium- and long-term follow-up, dietary interventions may result in little to no difference in BMI. One study (103 participants; low-certainty evidence) reported no difference in BMI at medium-term follow-up, and one study (2238 participants; low-certainty evidence) found lower rates of overweight and obesity in the intervention group at long-term follow-up. Dietary interventions likely result in little to no difference in zBMI at short-term (MD 0.03, 95% CI -0.13 to 0.19; 2 studies, 145 participants) and medium-term follow-up (MD -0.17, 95% CI -0.44 to 0.10; 1 study, 389 participants), both with moderate-certainty evidence. None of the included studies reported zBMI at long-term follow-up. None of the studies reported serious adverse events. Activity interventions versus control Activity interventions may have little to no effect on BMI at short-term follow-up, but the evidence is very uncertain (MD -0.10, 95% CI -0.28 to 0.08; 6 studies, 826 participants; very low-certainty evidence). They likely reduce BMI at medium-term follow-up (MD -0.70, 95% CI -1.09 to -0.31; 1 study, 567 participants; moderate-certainty evidence). None of the studies reported BMI at long-term follow-up. Activity interventions likely result in little to no difference in zBMI at short-term follow-up (MD -0.06, 95% CI -0.19 to 0.07; 3 studies, 635 participants; moderate-certainty evidence). They may result in little to no difference in zBMI at medium-term follow-up, but the evidence is very uncertain (MD -0.00, 95% CI -0.12 to 0.11; 4 studies, 1083 participants; very low-certainty evidence). None of the included studies reported zBMI at long-term follow-up. Activity interventions may have little to no effect on serious adverse events, but the evidence is very uncertain (2 studies, 773 participants; very low-certainty evidence). One study found no harms related to the intervention, and one study reported no difference in accident and infection rates between groups. Combined dietary/activity interventions versus control Combined dietary/activity interventions may have little to no effect on BMI at short-term follow-up, but the evidence is very uncertain (MD -0.08, 95% CI -0.20 to 0.04; 13 studies, 3867 participants; very low-certainty evidence). They may result in little to no difference in BMI at medium-term follow-up (MD -0.05, 95% CI -0.18 to 0.08; 9 studies, 7016 participants; low-certainty evidence), and may result in a slight reduction in BMI at long-term follow-up (MD -0.20, 95% CI -0.39 to -0.01; 5 studies, 2074 participants; low-certainty evidence). Combined interventions may have little to no effect on zBMI at short-term follow-up (MD -0.03, 95% CI -0.07 to 0.01; 14 studies, 5518 participants), and may result in a slight reduction in zBMI at medium-term follow-up, but the evidence for both time frames is very uncertain (MD -0.07, 95% CI -0.11 to -0.04; 15 studies; 11,043 participants). Combined interventions may result in a slight reduction in zBMI at long-term follow-up (MD -0.07, 95% CI -0.13 to -0.01; 10 studies, 4693 participants; low-certainty evidence). Combined interventions may result in little to no difference in serious adverse events, but the evidence is very uncertain (4 studies, 1689 participants; very low-certainty evidence). One study reported that a parent fractured an ankle while roller-skating at a community centre; the remaining studies reported no adverse events. AUTHORS' CONCLUSIONS: In early childhood, combined dietary/activity interventions may have very modest benefits on BMI and zBMI at long-term follow-up. When implemented alone, dietary or activity interventions may have little to no effect on BMI measures. Only six studies reported serious adverse events, with no serious harms resulting directly from the intervention, but the evidence is very uncertain. FUNDING: This review was partly funded by the National Institute for Health Research, School for Public Health Research. REGISTRATION: Protocol available: DOI: 10.1002/14651858.CD015326.