The objective of this study was to investigate the effects of chitosan (CHI) on ruminal and cecal fermentation and on apparent digestibility in sheep. In a first trial, 4 ruminally fistulated sheep fed an alfalfa hay and concentrate diet formulated to meet 1.2 times their maintenance energy requirements were assigned to 2 treatments [no-additive control (CTR), or 136 mg/kg of BW of CHI] for 19 d in a 2 x 2 crossover design. Samples of ruminal fluid and feces were taken for VFA and NH(3)-N concentrations. At the end of each period of the crossover, ruminal fluid was used as inoculum for batch cultures (24 and 96 h), and the effects on rate and extent of gas production and on fermentation variables for starch, cellulose, and the same diet fed to the animals were calculated. In a second trial, 4 sheep fed the alfalfa hay and concentrate diet were assigned to 2 treatments (CTR and CHI), as for the first trial, but this time for 14 d in a 2 x 2 crossover design. Apparent digestibility was studied. In the first trial, in terms of ruminal fermentation, CHI did not affect (P = 0.12) total VFA concentration or acetate proportion, but increased propionate proportion (P = 0.004) and propionate-to-acetate ratio (C3:C2; P = 0.007) and decreased branched-chain VFA (BCVFA) proportion (P = 0.003) and NH(3)-N concentration (P = 0.045). In feces, CHI decreased total VFA concentration (P = 0.01) and acetate (P = 0.045) and butyrate (P = 0.025) proportions, and increased propionate (P = 0.04), valerate (P = 0.01), and BCVFA (P = 0.032) proportions and C3:C2 ratio (P = 0.045). In batch cultures, when starch was used as substrate, CHI decreased acetate proportion (P = 0.016) and methane production (P = 0.03) and increased propionate and valerate proportions and C3:C2 ratio (P < 0.001). When cellulose was used as substrate, CHI decreased in vitro OM digestibility (P = 0.033), total VFA concentration, propionate proportion (P < 0.001), C3:C2 ratio (P = 0.006), methane production (P < 0.001), NH(3)-N concentration (P = 0.019), and gas production rate (P = 0.002), and increased butyrate (P = 0.029) and BCVFA (P = 0.004) proportions and pH (P < 0.001). When the alfalfa hay and concentrate diet was used as substrate, CHI decreased butyrate (P < 0.001) and BCVFA (P = 0.008) proportions, methane (P = 0.005) and asymptotic gas production (P = 0.02), and gas production rate (P = 0.05). In the second trial, CHI decreased NDF apparent digestibility (P = 0.09). In conclusion, CHI shifted ruminal fermentation toward energetically more efficient routes, without reducing OM apparent digestibility.