The great biochemical variability of the raw materials used in horse feeds can induce changes in gastric fermentation, and therefore affect gastric health through the resulting acidity of the bacterial end products. This in vitro trial compared the effect of 6 raw materials (BA: barley; WB: wheat bran; OA: oats; SM: sunflower meal; SB: sugar beet pulp; AL: alfalfa – Table 1) commonly used in horse feeds on gastric bacterial fermentation and buffering capacity. In 3 horses fed a mixture of the 6 raw materials 3h before sampling, gastric chyme was collected via a nasogastric tube and mixed. Bottles containing 1mL of the mixed gastric chymes, 50mL dilution medium and 1.0g of substrate (50% pilled barley + 50% one pilled raw material) were incubated in a shaking water bath at 39°C for 10h. Gas production, volatile fatty acids (VFAs), pH and buffering capacity of the solution (HCl 0.1M volume added to lower pH to 4 in the 50mL solution) were measured every 2h in duplicate. Effect of the substrate, time and their interaction were analyzed using a mixed procedure (SAS). From 4h to 10h incubation, gas production was greater with WB than BA, SM and AL (P < 0.001), reflecting higher bacterial fermentations, probably associated with WB highest soluble sugars content. Higher VFAs concentration were observed with AL than OA and WB at 2h, than BA and SB at 8h, than all other substrates at 10h (P > 0.001). This intense VFAs production with AL suggested environmental conditions more adapted for starch and soluble sugars fermentation, maybe combined with protein fermentation. Little differences were observed between raw materials regarding pH evolution. However, all along the 10h, buffering capacity of the solutions containing AL remained significantly higher than all others (P = 0.002, Table 1). The inclusion of alfalfa in horse feeds could be of interest to delay the decrease in gastric pH associated with HCl secretions and fermentation end products appearance, and to limit the exposure time of gastric mucosa to pH ≤ 4.