Abstract: Meat meal is a crucial animal protein feed ingredient, rich in high-quality proteins, fats, minerals, and vitamins, providing essential amino acids necessary for animal growth and immune function, thus exhibiting high nutritional value. However, during processing, transportation, and storage, meat meal is susceptible to environmental factors and storage conditions, leading to varying degrees of freshness and quality deterioration. During storage, the freshness of meat meal declines significantly over time due to microbial activity and enzymatic reactions, resulting in protein degradation and lipid oxidation, which cause nutrient loss. Concurrently, aldehydes, ketones, heterocycles, and acids accumulate as volatile organic compounds (VOCs), leading to undesirable odors such as rancidity and putrefaction, ultimately lowering the freshness grade of the feed ingredient. To analyze VOCs in meat meal at different freshness levels and explore the correlation between freshness indicators and VOCs, gas sensors responsive to freshness-related VOCs were selected. This study investigated pork meal and chicken meal as feed ingredients, employing standard methods to determine freshness indicators. Headspace solid-phase microextraction combined with gas chromatography-mass spectrometry (HS-SPME-GC-MS) was utilized to detect VOCs in meat meal samples with varying freshness levels. Extraction conditions were optimized using single-factor experiments and response surface methodology, determining the optimal parameters as follows: fiber type, 50/30 µm DVB/CAR/PDMS; extraction temperature, 80 ℃; extraction time, 40 min; equilibrium time, 30 min; and desorption time, 5 min. Based on the optimal extraction conditions, VOCs were analyzed in meat meal samples with different freshness levels, identifying 128 and 114 VOCs in pork meal and chicken meal, respectively. These VOCs encompassed alcohols, aldehydes, esters, ketones, acids, phenols, hydrocarbons, heterocycles, and others. Aldehydes (relative content: PM 14.5%～23.2%, CM 21.3%～36.8%) and heterocycles (relative content: PM 21.4%～36%, CM 6%～12%) were the predominant components. Freshness classification was performed based on four freshness indicators (TVB-N, AV, pH, and TVC). Orthogonal partial least squares discriminant analysis (OPLS-DA) was conducted using the relative abundances of 242 VOCs obtained via GC-MS as input variables (X), while freshness classification results were used as categorical labels (Y) to establish the OPLS-DA model. The model results demonstrated robust predictive performance, with R²X = 0.872, R²Y = 0.982, and Q² = 0.941 for PM, and R²X = 0.819, R²Y = 0.987, and Q² = 0.965 for CM, both satisfying Q² > 0.5. Furthermore, variable importance in projection (VIP) scores were used to identify key VOCs influencing meat meal freshness, with variables having VIP > 1 considered the most influential. Based on this criterion, 38 and 31 VOCs were identified as key freshness- and quality-related compounds in pork meal and chicken meal, respectively, with 10 shared key VOCs: three hydrocarbons (2,6,10-trimethyl tetradecane, heptadecane, and naphthalene), four aldehydes (benzaldehyde, 2-butyl-2-octenal, nonanal, and (Z)-hexadecenal), one acid (hexanoic acid), one ketone (2-decanone), and one alcohol (1-octen-3-ol). To further elucidate the relationship between key VOCs and meat meal freshness, Pearson correlation analysis was performed between freshness indicators and VOCs, identifying 25 and 17 VOCs in pork meal and chicken meal, respectively, as significantly correlated with freshness indicators (TVB-N 、AV、pH、TVC). Among these, heptadecane, 1-octen-3-ol, 2-butyl-2-octenal, and (Z)-hexadecenal were shared characteristic VOCs. Based on the selected characteristic VOCs, 10 gas sensors potentially responsive to freshness indicators were chosen. These findings provide a scientific basis for VOC detection and characterization in meat meal, facilitating the selection of gas sensors for electronic nose-based freshness detection of meat meals.