Fodder maize is the most commonly used crop for biogas production owing to its high yields, high concentrations of starch and good digestibility. However, environmental concerns and possible future conflict with land for food production may limit its long-term use. The bioenergy grass, Miscanthus, is a high-yielding perennial that can grow on marginal land and, with ‘greener’ environmental credentials, may offer an alternative. To compete with maize, the concentration of non-structural carbohydrates (NSC) and digestibility may need to be improved. Non-structural carbohydrates were quantified in 38 diverse genotypes of Miscanthus in green-cut biomass in July and October. The aim was to determine whether NSC abundance could be a target for breeding programmes or whether genotypes already exist that could rival maize for use in anaerobic digestion systems. The saccharification potential and measures of N P and K were also studied. The highest concentrations of NSC were in July, reaching a maximum of 20% DW. However, the maximum yield was in October with 300–400 g NSC plant
−1 owing to higher biomass. The digestibility of the cell wall was higher in July than in October, but the increase in biomass meant yields of digestible sugars were still higher in October. Nutrient concentrations were at least twofold higher in July compared to November, and the abundance of potassium showed the greatest degree of variation between genotypes. The projected maximum yield of NSC was 1.3 t ha
−1 with significant variation to target for breeding. Starch accumulated in the highest concentrations and continued to increase into autumn in some genotypes. Therefore, starch, rather than sugars, would be a better target for breeding improvement. If harvest date was brought forward to autumn, nutrient losses in non-flowering genotypes would be comparable to an early spring harvest.