Abstract: This study developed a electrochemical sensor based on gold nanoparticles (AuNPs) and multi-walled carbon nanotubes (MWCNTs) modified anodized aluminum oxide (AAO) membrane, coupled with a screen-printed carbon electrode (SPCE), for trace cadmium (Cd²⁺) and lead (Pb²⁺) ion detection in grain samples. The surface morphology and structure of the AuNPs-MWCNTs@AAO membrane were characterized by scanning electron microscopy (SEM), transmission electron microscopy (TEM), and X-ray diffraction (XRD). Electrochemical performance was assessed using cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS). The optimized detection conditions were as follows: AAO membrane ultrasound treatment for 10 minutes, buffer solution at pH5.5, and stripping at −1.2 V for 180 s in the optimized buffer. Over the concentration range of 0.1~10.00 μg/L, the oxidation peak currents exhibited a strong linear correlation with concentration, with correlation coefficients (R²) of 0.9903 and 0.9993. The limits of detection (LOD) for Cd²⁺ and Pb²⁺ were 0.23 μg/L and 0.11 μg/L, respectively. The sensor demonstrated strong selectivity against common interfering metal ions, with current response variation not exceeding 5% even at 100-fold excess concentrations. The sensor exhibited high repeatability, reproducibility, and long-term stability, with relative standard deviations (RSD) of 2.5% and 2.3% for the oxidation-reduction peak currents. The sensor is easy to prepare and provides reliable, effective performance for detecting trace heavy metals in grain samples.