Abstract: To find the rapid method to detect residual?antibiotic in dairy products, discrimination on the same concentration of six kinds of antibiotics in milk powder using smartongue and detection on the concentration of neomycin is researched in this paper. Platinum, gold, palladium, tungsten, titanium and silver electrodes as sensor array were used in this experiment and detected at different frequency of 1 Hz, 10 Hz and 100 Hz. Principal component analysis, linear discriminant analysis and partial least square method were used to analysis the experiment data. The results showed that smartongue could well discriminate different antibiotic and different concentration of neomycin, qualitative analysis result could reach the national maximum residue limit standard basically. A optimal concentration range of neomycin about 300-1?100?μg/L was obtained by quantitive model established by partial least square method. Smartongue provide a new method to detect adulterations in food by its special advantages.