Background: Childhood asthma has been linked to distinct metabolomic profiles. Objective: We sought to identify phenotypes (metabotypes) in children with moderate to severe asthma through integrative fecal and serum metabolome analysis. Methods: Children from the Systems Pharmacology Approach to Uncontrolled Pediatric Asthma cohort with Global Initiative for Asthma treatment step 3 or higher were recruited. Asthma control was defined by the Asthma Control Test and annual exacerbation history. Targeted metabolomic profiling of feces and serum was performed using liquid chromatography and flow injection electrospray ionization-triple quadrupole mass spectrometry. Similarity network fusion integrated fecal and serum metabolome profiles, followed by spectral clustering. Clusters were analyzed for differences in asthma characteristics, food diaries, fecal microbiota composition, and levels of serum inflammatory markers and blood cells. Results: Integrative fecal and serum metabolome analysis of 92 children with moderate to severe asthma (median age, 11.5 years, 34% female) revealed 3 metabotypes. Metabotype 1 had the lowest percentage of allergic rhinitis, with elevated serum ceramides and triglycerides. Metabotype 2 had higher odds of asthma control, the highest percentage of children with 4 or more months of breast-feeding, reduced sugar intake, lowest levels of blood neutrophils and serum inflammatory markers, and elevated serum acylcarnitines and ω-3 fatty acids. Metabotype 3 included the highest percentage of uncontrolled asthma patients, with decreased serum cholesteryl esters, phosphatidylcholines, and sphingomyelins, elevated fecal amino acids, and reduced fecal microbiota diversity. Conclusions: Metabotypes in children with moderate to severe asthma are linked to asthma control, distinct fecal microbiota, and systemic inflammatory patterns. The findings suggest that metabotyping can be valuable in precision medicine approaches for asthma.