Seven feed additives (i.e., quillaja saponins, fumaric acid, garlic oil, fish oil, cinnamaldehyde, monensin, and medium chain fatty acids (MCFA)), were evaluated for their effects on CH4 inhibition in vitro in combination with four substrates: concentrate (CON), grass silage (GS), maize silage (MS) and the mixture of CON + GS + MS (300:350:350 (dry matter, DM), MIX), all feeds regularly used in dairy cattle feeding. Substrates and additives were incubated in a batch incubation system containing buffered rumen fluid for 24 h. Cinnamaldehyde had an interaction with substrate for CH4 inhibition, volatile fatty acid (VFA) production and inhibition of CH4 relative to VFA. For fumaric acid, interactions occurred for CH4 relative to VFA. Fish oil, quillaja saponins and MCFA had additive x substrate interactions for inhibition of CH4 and VFA production, but they had no interactions for CH4 relative to VFA. Garlic oil had no interaction with substrates for CH4 production and CH4 relative to VFA, but had interactions for VFA production. Monensin had substrate x additive interactions for CH4 and VFA production and CH4 relative to VFA. Monensin and quillaja saponins were more effective at inhibiting CH4 production when combined with GS and MS than with CON. Fish oil had higher inhibition when combined with CON and GS than with MS. The MCFA had higher inhibitions when combined with MS and were lowest with CON, the combination with GS was intermediate and differed from both other substrates. Cinnamaldehyde and MCFA decreased, whereas fumaric acid increased the total VFA production. No other additive affected total VFA production. As a general CH4 mitigation strategy, fumaric acid, garlic oil and fish oil were better in combination with CON. Monensin was more effective in combination with GS, and quillaja saponins were more effective when combined with MS. Cinnamaldehyde and MCFA strongly inhibited fermentation which impaired appropriate evaluation of the most promising combination. Despite additive x substrate interactions, CH4 and VFA production in incubations with CON, GS and MS did not differ from the weighted average of incubations with single substrates with or without additives. Hence no synergism between additive and substrate combinations seemed to exist. Results clearly indicate interactions between additives and substrates. However, it is unlikely that this interaction is the origin of often variable results among in vitro studies. (C) 2012 Elsevier B.V. All rights reserved.