The effect of slurry type (pig or cow) and use of a bacterial silage inoculant, was assessed on herbage from long-term slurry plots ensiled in pilot-scale (0·7 tonne) silos. Silage was fed to sheep to assess digestibility and metabolisable energy content. The latter was measured by direct calorimetry in a modified cattle respiration chamber. Non-inoculated slurry-treated silages had poorer fermentation quality than inoculant-treated silages (P < 0·001) and silages treated with cow slurry had poorer fermentation quality than those treated with pig slurry (P < 0·001). Digestibility values for different components of the silages showed few differences. Energy digestibility values of pig slurry-treated silages were lower (P < 0·05) than for cow slurry-treated silages. Methane production from all animals in the experiment was comparatively low at 5·97–6·74% of gross energy intake. Metabolisable energy contents of the cow slurry-treated silages were higher than pig slurry-treated silages (P < 0·05) and the ME contents of inoculant treated silages were higher than of non-inoculated silages (P < 0·001). Thus, treatment of slurry-treated silages with bacterial inoculant enhanced fermentation quality and ME contents. The effect of slurry type on fermentation quality and feeding value was ambiguous; cow slurry-treated silages exhibiting poorer fermentation quality but higher ME contents than equivalent pig slurry-treated silages. © 1998 Society of Chemical Industry.