Small intestinal bacterial overgrowth (SIBO) is highly prevalent and is associated with numerous gastrointestinal disorders, but the microbes involved remain poorly defined. Moreover, existing studies of microbiome alterations in SIBO have utilized stool samples, which are not representative of the entire gastrointestinal tract. Therefore, we aimed to determine and compare the duodenal microbiome composition in SIBO and non-SIBO subjects, using duodenal aspirates from subjects undergoing standard-of-care esophagogastroduodenoscopy without colon preparation. Using the recently-redefined cutoff for SIBO of >103 colony forming units per milliliter (CFU/mL), 42 SIBO and 98 non-SIBO subjects were identified. Duodenal samples from SIBO subjects had 4x103-fold higher counts than non-SIBO subjects when plated on MacConkey agar (P<0.0001), and 3.8-fold higher counts when plated on blood agar (P<0.0001). Twenty subjects had also undergone lactulose hydrogen breath tests (LHBTs), of whom 7/20 had SIBO. At the 90-minute timepoint, 4/7 SIBO subjects had positive LHBTs (rise in hydrogen (H2) ≥ 20 ppm above baseline), as compared to 2/13 non-SIBO subjects. 16S ribosomal RNA (rRNA) sequencing revealed that SIBO subjects had 4.31-fold higher relative abundance of Proteobacteria (FDR P<0.0001) and 1.64-fold lower Firmicutes (P<0.0003) than non-SIBO subjects. This increased relative abundance of Proteobacteria correlated with decreased α-diversity in SIBO subjects (Spearman R = 0.4866, P<0.0001) Specific increases in class Gammaproteobacteria correlated with the area-under-the-curve for H2 for 0-90 mins during LHBT (R = 0.630, P = 0.002). Increases in Gammaproteobacteria resulted primarily from higher relative abundances of the family Enterobacteriaceae (FDR P<0.0001), which correlated with the symptom of bloating (Spearman R = 0.185, 2-tailed P = 0.028). Increases in family Aeromonadaceae correlated with urgency with bowel movement (Spearman R = 0.186, 2-tailed P = 0.028). These results validate the >103 CFU/mL cutoff for the definition of SIBO, and also reveal specific overgrowth of Proteobacteria in SIBO vs. non-SIBO subjects, coupled with an altered Proteobacterial profile that correlates with symptom severity. Future research may elucidate host-microbiome interactions underlying these symptoms in SIBO patients.