In vitro methods have been developed to measure digestibility, but such methods may not accurately estimate methane or volatile fatty acid (VFA) production. Methane emissions are stoichiometrically linked with VFA profiles. For example, a shift from acetate to propionate may decrease CO₂ and H₂ production, and in turn decrease conversion of CO₂ and H₂ to methane. The objective of this study was to determine the effect of different conditions of in vitro procedures on VFA and gas profiles in comparison with in vivo measurements. Experimental design was a 4 × 2 × 2 factorial CRD with 4 replicates. Treatments were 4 ratios of rumen fluid to buffer by volume (95/5, 75/25, 50/50, 25/75), 2 concentrations (w/v) of added timothy hay (0.5% or 1%), and with or without sodium acetate addition (50 mmol final concentration). Statistical analysis was conducted using a mixed model that included all fixed effects and interactions. Total volume of broth (rumen fluid and buffer) was 10 mL per tube. Measurements of gas production and VFAs were recorded at 0, 4, 16, 24, and 48 h. Total gas was proportioned into CO₂ and non-CO₂ after collection at 48 h. Higher hay concentration averaged more (P < 0.0001) total gas than the lower hay concentration (12.1 vs. 6.7 mL gas/tube; SE = ± 0.23), and more (P < 0.0001) non-CO₂ gas (0.28 vs. 0.15 mL; SE = ± 0.046). Total gas production increased (P < 0.001) with higher buffer concentration, and averaged 6.1, 9.1, 11.2 and 11.2 mL per tube (SE = ± 0.32 mL) as buffer concentration increased. The acetate/propionate (A/P) concentration decreased over time (P < 0.0001). The initial rumen fluid A/P ratio was 3.7 but the A/P ratio of produced VFA averaged 2.5 (SE = ± 0.14). The addition of acetate did not affect gas production or A/P ratio of produced VFA. This could mean that adding acetate to a system does not necessarily shift production away from acetate. A/P ratio differed for VFA produced in vitro compared with initial rumen fluid, but no tested treatments were identified as a cause of the difference.