NiO, CuO, and Co3O4 nanoparticles supported by carbon nanotubes (CNTs) were synthesized for the electrochemical detection and conversion of glucose. The structures of the as-prepared catalysts were characterized by transmission electron microscopy, X-ray diffraction, Raman spectroscopy, and inductively coupled plasma-optical emission spectrometry. The electrochemical performance of the catalysts was evaluated by cyclic voltammetry, linear scanning voltammetry, electrochemical impedance spectroscopy, and chronoamperometry measurements. The optimal potentials of the resulting sensors were determined from the current response as obtained from the current–time tests. The minimum limit of detection (LOD), stability, reproducibility, and interference resistance of the sensors were evaluated. The resulting electrochemical sensors showed excellent sensitivity, stability, reproducibility, and interference resistance for the detection of glucose. In particular, the electrochemical sensor based on Co3O4/CNTs had an LOD of 2 μM in glucose concentrations ranging from 0.004 to 0.474 mM. The products of glucose oxidation at potentials of 1.466, 1.666, and 1.866 V with varying electrolysis times of 1, 2, 4, and 6 h, respectively, were analyzed using high-performance liquid chromatography. Five compounds including glucuronic acid (GNA), glucuronic acid, formic acid, oxalic acid, and ethanoic acid were detected. After 4 h electrolysis at 1.666 V, the glucose conversions resulting from NiO/CNTs, CuO/CNTs, and Co3O4/CNTs were 56.5, 43.2, and 53.9%, respectively. Notably, Co3O4/CNTs yielded the highest GNA selectivity of 46.8%. In contrast, NiO/CNTs and CuO/CNTs tended to produce more C1 and C2 compounds.