Sound-absorbing materials have been widely used to reduce noise pollution. Waste lignocellulose can be expected to prepare sound-absorbing boards, due to the environmental benefit and low price. Among them, one of the lignocellulosic wastes, sugarcane bagasse (SCB) can be conveniently collected and transported from sugar factories. SCB can be either directly compressed into sound-absorbing boards with the rest of the materials, such as polyurethane foam or red clay, to form composite sound-absorbing boards, or modified and then mixed with materials, like bamboo charcoal, to produce composite sound-absorbing boards. However, it is still lacking in the effect of chemical treatment on the sound absorption performance of SCB. Taking alkaline reagents to treat SCB, this study aims to investigate the effects of alkaline treatment on the sound absorption performance of SCB boards. Sound-absorbing boards were then optimized with alkali-treated SCB after preparation. Firstly, Box-Benhnken design was applied to optimize the parameters during alkaline treatment, including liquid-solid ratio (10:1, 15:1, and 20:1), NaOH concentration (0.5%, 2.25%, and 4%), reaction temperature (55, 70 and 85 °C), and reaction time (0.5, 1.5 and 2.5 h) with sound absorption coefficient (SAC) as the function. The NaOH treatment improved the SAC of SCB at the medium (1 000-4 000 Hz) and high (4 000-6 300 Hz) frequencies, but there was a decrease at low (500-1 000 Hz) frequencies. Quadratic surface response method (SRM) was established at 1 000-4 000 Hz and 4 000-6 300 Hz. The suboptimal prediction was attributed to the influence of board-making conditions during sound absorption coefficient testing. The structure and morphology of SCB were two of the influencing factors on the SAC after alkaline treatment. The SCB was treated at the liquid-to-solid ratio of 15:1, 4% NaOH, 70 °C for 0.5 h, due to the highest SAC at 4 000-6 300 Hz. Physic-chemical characterization was selected, including compositional analysis via two-step acid hydrolysis, specific surface area and pore size by Brunauer-Emmett-Teller (BET), and surface morphology via scanning electron microscope (SEM) and atomic force microscope (AFM). The results showed that the SAC was related to the compositional and surface roughness of SCB after alkali treatment. The SCB after NaOH treatment contained the a higher content of cellulose with the layered structure of linear macromolecular chains and abundant hydroxyl groups, leading to the high frictional loss and energy conversion efficiency of medium and high-frequency sound waves. The rougher surface of NaOH-treated SCB also improved SAC at 1 000-4 000 and 4 000-6 300 Hz via reflection, scattering, and energy conversion of sound waves. There was an increase in the specific surface area, pore size, and total pore volume of alkali-treated SCB. The hot-pressing pressure also caused the compression and collapse of the large-pore-size/volume structure after alkali treatment. Phenolic resin adhesive was used to fill the pore channels during gluing. Thereby, the SAC was offset after modification of the specific surface area and pore size. As a result, the porous characteristics of bagasse samples failed to effectively reflect the sound absorption performance of the boards. Therefore, the material was selected as alkali-treated SCB with the best sound absorption performance at 4 000-6 300 Hz. Orthogonal test was conducted to optimize the board preparation, including hot pressing temperature (110, 130, and 150 °C), hot pressing pressure (2.5, 5 and 7.5 MPa), hot pressing time (5, 10, and 15 min) and glue amount (12%, 15%, and 18%). The optimal temperature of the hot pressing was obtained to improve the average SAC. The hot compression pressure shared had various influences on the average SAC at different frequencies. The hot-pressing time and glue amount also improved the average SAC. The optimal preparation conditions of SCB boards with better SAC were obtained as follows: hot pressing temperature, pressure, time, and glue amount were 130 °C, 7.5 MPa, 15 min, and 18%, respectively. The average SAC values of the SCB boards with a thickness of 5 mm were 0.376 and 0.909 at the frequencies of 1000-4000 Hz and 4000-6300 Hz, respectively. This finding can provide the synergistic impact of the material properties and board-making conditions on the sound absorption performance of boards with lignocelluloses.