Abstract Context and Objective The impact of existing malnutrition on stroke outcomes is poorly recognised and treated. Evidence was systematically reviewed and quantified by meta-analysis. Methods MEDLINE, EMBASE and Web of Science were searched from inception to 11 January 2021 and updated in July. Prospective cohort studies, in English, evaluating anthropometric and biomarkers of nutrition on stroke outcomes were included. Risk of bias was assessed using the Scottish Intercollegiate Guidelines Network checklist. Results Twenty-six studies (n = 156 249) were eligible (follow-up: One month-14 years). Underweight patients had increased risk of long-term mortality (adjusted hazard ratio = 1.65,1.41-1.95), whilst overweight (0.80,0.74-0.86) and obese patients (0.80,0.75-0.85) had decreased risk compared to normal weight. Odds of mortality decreased in those with high serum albumin (odds ratio = 0.29,0.18-0.48) and increased with low serum albumin (odds ratio = 3.46,1.78-6.74) compared to normal serum albumin (30-35 g/L). Being malnourished compared to well-nourished, as assessed by the Subjective Global Assessment (SGA) or by a combination of anthropometric and biochemical markers increased all-cause mortality (odds ratio = 2.38,1.85-3.06) and poor functional status (adjusted odds ratio = 2.21,1.40-3.49). Conclusion Nutritional status at the time of stroke predicts adverse stroke outcomes.