In 2017, the FDA approved for the first time in 20 years a new drug to treat sickle cell disease (SCD). The drug, Endari, is an oral L-glutamine metabolite that acts by reducing oxidative stress in erythrocytes. The conclusions from the phase 3 randomized clinical trial suggested that treatment with L-glutamine reduces pain crises and hospitalization rate (Niihara et al., NEJM, 2018). Our goals were to confirm the clinical trial findings using an alternative, genetics-based strategy, and to identify other metabolites causally implicated in the clinical heterogeneity observed in SCD patients. Employing a targeted metabolomics platform, we profiled 129 metabolites from the plasma of two SCD cohorts from Paris, France (GEN-MOD (n=406)) and North Carolina, USA (OMG-SCD (n=300)). To prioritize metabolites, we performed correlation analyses between metabolite levels and 6 SCD complications (pain crises, cholecystectomy, retinopathy, leg ulcer, priapism, aseptic necrosis) or estimated glomerular filtration rate (eGFR).We assessed statistical significance via permutation test and used 2-sample Mendelian randomization (MR) methods to test for causality. We obtained MR instruments associated with metabolites of interest from two large metabolite GWAS performed in healthy Europeans (Shin et al., Nature Genetics, 2014, n=7,824; Long et al., Nature Genetics, 2017, n=1,960). We derived SNP associations with SCD clinical outcomes from the large African-American CSSCD cohort (n=1,278). We identified a causal relationship between L-glutamine levels and pain crises requiring hospitalization (odds ratio (OR) [95% confidence interval]=0.68 [0.52 – 0.89], P=0.005). Our MR study estimates a 32% reduction in pain crises with one standard deviation increase of L-glutamine. This estimate is comparable to the 33% reduction in hospitalization due to emergency room visit observed in SCD patients treated with L-glutamine vs. placebo in the clinical trial. We identified 71 other significant associations (FDR<5%) between metabolite levels and SCD-related complications or eGFR. Two-sample MR analyses revealed a causal association between 3-ureidopropionate and eGFR: a standard deviation increase in 3-ureidopropionate leads to an eGFR increase of 0.07 mL/min per 1.172 m2 (P=9.7x10-4). MR analyses provide additional evidence to support the causal role of L-glutamine in modulating pain crises in SCD. Additionally, we found a tentative causal link between 3-ureidopropionate levels and eGFR.