Agricultural biogas sector is currently facing fast development in France, and cattle manure which is rich in straw constitutes an important feedstock available in an agricultural context. Cattle manure is characterized by a total solid content of about 20-30 %, and it is thus suited to dry anaerobic digestion. Among the available technologies, batch dry digesters processes with leachate spreading are particularly interesting for small-scale biogas plants. However, the start-up of these plants is quite slow in reason of low hydrolysis rate of cattle manure, and codigestion represents a good option to facilitate the start-up phase. In this study, sorghum and miscanthus were considered for codigestion with manure, because these energy crops present the advantages of high biomass yields and low inputs requirements (water, fertilizers). In addition miscanthus can grow on polluted soils whereas sorghum can be cultivated as catch crops. Nevertheless carbohydrates accessibility of these lignocellulosic biomasses is limited and may be improved before anaerobic digestion processes by alkali pretreatments. As a consequence, the objective of this study was to evaluate the efficiency of alkali pretreatments of sorghum and miscanthus to improve the start-up phase and the reactor performance when codigested with cattle manure. Experiments were carried out in 6 L leach bed reactors (LBR), fed with 300 g TS of organic substrate and 1.1 L leachate. The substrate/inoculum ratio was 6 (gVS/gVS) by adding a digestate sampled in a previous batch, and the substrate was composed of 85% (in wet weight basis) of manure and 15% of sorghum or miscanthus (corresponding to 39-40%VS of sorghum or miscanthus). Two different pretreatments were applied at room temperature, with low water addition (to reach 13%TS) and no mixing: soda (NaOH, 10g/100 gTS, 1 day) pretreatment for sorghum only, lime pretreatment (10g/100 gTS, for 5 days) for sorghum and miscanthus. The pretreated substrates as well as the raw ones without pretreatment (controls) were digested in LBR in duplicates during two different runs. After 57 days of sorghum codigestion, the highest methane production was obtained for NaOH- pretreatment with a methane yield of 232 ± 15 mL/gVS, corresponding to an increase of 20 % and 11% compared to the control and CaO-treated sorghum, respectively. However, during the start-up phase of the batch process, soda pretreatment was less efficient than lime pretreatment because, within 6 first days, the highest methane production (61 ± 1 mL/gVS) was obtained for lime-pretreated sorghum (43 ± 22 mL/gVS and 25 ± 16 mL/gVS for soda-pretreated sorghum and untreated sorghum, respectively). In fact, soda pretreatment led to a transient acidification phase with an accumulation of volatile fatty acid (VFA) concentration up to 9.3 g/L on days 2-3, which slowed down the methanogenic activity. In contrast, maximum VFA concentrations were 6.8 g/L and 6.5 g/L for lime pretreated sorghum and raw sorghum, respectively. Leachate pH did not fall under 6.48 thanks to the presence of alkali, which may have mitigated the negative effect of VFA accumulation. Concerning the codigestion of miscanthus with cattle manure, lower methane production than codigestion with sorghum was observed: 158 ± 4 mL/gVS after 59 days for untreated miscanthus. Lime pretreatement led to only a 6 % increase in methane production. A slightly higher impact (+8%) was observed on methane production after 6 days. Lower accumulation of VFA was observed than for sorghum reactors (e.g., 5 g/L after 1 day for lime pretreated miscanthus). In conclusion, alkali pretreatment of sorghum and miscanthus before their batch codigestion with manure showed a slight positive impact on reactor performances. Soda pretreatment led to a higher acidification risk during the start-up phase. Associated with the detrimental impact of sodium on agricultural soils in the case of digestate land spreading, lime pretreatment is recommended.