Objective: To develop a multiplex visual meat adulteration detection method, dual-tailed recombinase polymerase amplification (RPA) with nucleic acid hybridization-lateral flow strip (NAH-LFS) was established for the simultaneous identification of chicken, pork, and horse derived components in donkey meat and its products. This RPA-NAH-LFS method met the requirements of meat adulteration detection for rapidity, simplicity, and low cost. Methods: Firstly, samples was prepared with different adulteration ratios and perform DNA extraction. The primer5 software was utilized to design specific primers targeting the mitochondrial cytochrome b (cytb) gene of chicken, pork, and horse. The primers included three functional regions: primer sequence, spacer9, and single-stranded DNA tag sequence (as a tailing sequence). Spacer9 could prevent the action of DNA polymerase, allowing the single-stranded DNA tag sequence to remain single-stranded after RPA amplification. Subsequently, colloidal gold nanoparticles were prepared and conjugated with recognition probes. NAH-LFS were then constructed. The RPA reaction conditions and the optimal working parameters were optimized for the LFS, followed by systematic validation of the sensitivity and specificity of the RPA-NAH-LFS. Finally, the established assay was applied to analyze 40 commercial donkey meat samples. Results: The optimal RPA conditions were determined as follows: reaction temperature of 39 ℃, incubation time of 10 min, Mg²⁺ (280 mmol/L) addition volume of 2.5 μL, and primer addition volumes of 0.7 μL (chicken), 0.7 μL (pork), and 1.0 μL (horse). The NC membrane with a flow rate of 33.75 cm/s and the running buffer 4 (0.01 mol/L PBS, pH8.0, 0.05% Tween-20) were selected as optimal conditions. The LFS had three test lines targeting for chicken (T₁), pork (T₂), and horse (T₃). The detection process was completed within 15 min, which included RPA reaction (10 min) and LFS detection (5 min). The method could detect as little as 0.01% (w/w) chicken, pork, or horse meat in simulated adulterated donkey meat. Cross-reactivity test with 9 meat species demonstrated the excellent specificity of the assay. Analysis of the 40 commercial meat products revealed pork adulteration present in one sample, and simultaneous chicken and pork adulteration in another sample. These results were 100% consistent with those obtained by the standard detection method (DB42/T 1591-2020). Conclusion: The RPA-NAH-LFS technique, utilizing the dual-tailed primer design and a multiplex probe hybridization strategy, enables the simultaneous visual detection of meat adulterants. This method has high sensitivity, strong specificity, and operational convenience, making it particularly suitable for rapid on-site screening by basic-level regulatory agencies. It provides an efficient technology for detecting the authenticity of meat products.