Developing a sustainable biofuel program that minimizes negative impacts on the environment and captures the maximum usable energy per acre is a major challenge in the 21st century. This study used rice as a model grass to identify novel biomass genes and to determine natural variation in cell-wall composition given the wide array of rice genomic tools and the ease of genetic manipulation relative to emerging grass bioenergy crops. In rice, we established an efficient pipeline to identify loci for biomass production and composition. From that work, we identified biomass QTL and candidate genes for improving biomass quality and quantity in plants. Then, we developed resources to enhance the translation of information of biomass production from rice to switchgrass, and used these resources to test candidate genes for function in improving biomass production in switchgrass. We created a web based genomics platform to display and facilitate analysis of rice and switchgrass comparative functional and genomic data. Finally, we describe a novel field-level high throughput phenotyping system that allows for efficient, rapid capture of phenotype data, enabling the mapping of biomass QTL in large plant populations during growth and development. Overall, our project contributed (1) valuable genetic and genomic resources and tools in rice and switchgrass to the biomass community, (2) identified key biomass traits that are amenable for improvement by breeding or biotechnology, (3) provided validation that information from model systems could be applied to improvement of bioenergy crops, (4) developed a novel high throughput phenotyping system for field-level discovery of biomass QTL, and (5) contributed to the interdisciplinary training of four post doctoral fellows and five graduate students.