Obesity and adverse metabolic outcomes in patients with severe mental illness are clinically significant but potentially preventable. Importantly, the evidence for switching to antipsychotics to reduce cardiometabolic burden is unclear. PubMED, Embase, PsycINFO, and Cochrane were searched from inception to March 8, 2020. Articles reporting weight and metabolic changes after antipsychotic switching vs staying on the previous antipsychotic were meta-analyzed both across and within group. Of 61 identified studies, 59 were meta-analyzed (40% rated high quality). In the switch-vs-stay pairwise meta-analyses, only aripiprazole significantly reduced weight (-5.52 kg, 95% CI -10.63, -0.42, P = .03), while olanzapine significantly increased weight (2.46 kg, 95% CI 0.34, 4.57, P = .02). Switching to aripiprazole also significantly improved fasting glucose (-3.99 mg/dl, 95% CI -7.34, -0.64, P = .02) and triglycerides (-31.03 mg/dl, 95% CI -48.73, -13.34, P = .0001). Dropout and psychosis ratings did not differ between switch and stay groups for aripiprazole and olanzapine. In before-to-after switch meta-analyses, aripiprazole (-1.96 kg, 95% CI -3.07, -0.85, P < .001) and ziprasidone (-2.22 kg, 95% CI -3.84, -0.60, P = .007) were associated with weight loss, whereas olanzapine (2.71 kg, 95% CI 1.87, 3.55, P < .001), and clozapine (2.80 kg, 95% CI 0.26, 5.34, P = .03) were associated with weight gain. No significant weight or other cardiometabolic changes were observed when switching to amisulpride, paliperidone/risperidone, quetiapine, or lurasidone. Switching antipsychotics to agents with lower weight gain potential, notably to aripiprazole and ziprasidone, can improve weight profile and other cardiometabolic outcomes. When choosing switch agents, both the weight gain potential of the pre- and post-switch antipsychotic must be considered. Antipsychotic switching in psychiatrically stable patients must be weighed against the risk of psychiatric worsening.