Feeding plants containing elevated levels of polyphenols may reduce ruminal CH4 emissions, but at the expense of nutrient utilisation. There might, however, be non-additive effects when combining high-phenolic plants with well-digestible, high-nutrient feeds. To test whether non-additive effects exist, the leaves of Carica papaya (high in dietary quality, low in polyphenols), Clidemia hirta (high in hydrolysable tannins), Swietenia mahagoni (high in condensed tannins) and Eugenia aquea (high in non-tannin phenolics) were tested alone and in all possible mixtures (n 15 treatments). An amount of 200 mg DM of samples was incubated in vitro (24 h; 39oC) with buffered rumen fluid using the Hohenheim gas test apparatus. After the incubation, total gas production, CH4 concentration and fermentation profiles were determined. The levels of absolute CH4, and CH4:SCFA and CH4:total gas ratios were lower (P< 0·05) when incubating a combination of C. papaya and any high-phenolic plants (C. hirta, S. mahagoni and E. aquea) than when incubating C. papaya alone. Additionally, mixtures resulted in non-additive effects for all CH4-related parameters of the order of 2–15 % deviation from the expected value (P< 0·01). This means that, by combining these plants, CH4 in relation to the fermentative capacity was lower than that predicted when assuming the linearity of the effects. Similar non-additive effects of combining C. papaya with the other plants were found for NH3 concentrations but not for SCFA concentrations. In conclusion, using mixtures of high-quality plants and high-phenolic plants could be one approach to CH4 mitigation; however, this awaits in vivo confirmation.