Early seedling vigor and juvenile vegetative growth are important traits that allow the strong establishment of plants and access to nutrients and water, providing competition against weeds, and allowing mechanical cultivation in production systems that do not use herbicides. Drought stress at this early growth stage may be lethal or damaging. We used to the plant Digital Biomass as predicted from digital images to track plant growth under both well-watered and water-stressed conditions. To achieve these goals, we developed a manual imaging system that allowed us to track the plant growth over a period of 32 days. We imaged 30,36 plants representing 449 inbred lines daily from 13 to 32 days after planting with both a top and a side image. The drought treatment started 23 days after planting by completely withholding water from the water-stress treatment. Using Integrated Analysis Platform (IAP) software, we extracted 137 traits from the images including plant architectural traits and color traits. Phenotypic analysis of several traits showed variability across inbreds. Digital Biomass, for example, showed a great variability across inbreds with a 6.6-fold difference at the beginning of the experiment. Digital Biomass, estimated from the top and side images, was shown to be a good measure of plant vigor and strongly correlated with plant shoot weight at harvest. Vigorous seedling utilized more water, reflecting their ability to take advantage of available resources. The value of image-based traits of young plants was evaluated as a predictive tool for adult phenotypes grown in the field. Weak to moderate correlations were obtained between Digital Biomass at the seedling stage, with r-squared values of -0.35, -0.31 for GDD to Anthesis, and GDD to Silking respectively. The correlation between early maize growth and flowering time may suggest a common genetic control of growth and development of both stages with some possible genes with pleiotropic effects. To identify genomic regions associated with the several phenotypic traits, we utilized a dataset of 436,576 SNP markers to conduct Genome-wide Association (GWAS) using the GAPIT package in R. Several candidate genes were identified for growth rate and total leaf area at specific growth stages, as well as for other correlated traits. GWAS of image-derived plant color traits detected genes associated with plant pigments such as anthocyanin and chlorophyll, which confirms earlier reports on the utility of plant imaging in identifying plant pigments. We wanted to test whether growth, as measured by Digital Biomass, was controlled by a fixed or a dynamic set of genes, so we carried out GWAS analysis of Digital Biomass for each day as a separate phenotype. Results have shown that variation for early vegetative growth in maize is controlled by a dynamic set of genes over time, highlighting the importance of repeated measurement over time in GWAS and QTL studies designed to characterize the genetic architecture of plant development. The analysis of the drought-stressed plants showed variability in different drought tolerance traits ranging from 1.2 to 12.2-fold difference. The several measured traits included traits such as 1) leaf expansion sensitivity to water content and traits related to the ability to recover after drought such as 2) surviving green tissue after drought stress, 3) water use efficiency, and 4) growth rate after recovery with. No or weak correlations were found between the plant's ability to tolerate drought and its ability to recover. Photosynthesis Efficiency measured as Fv/Fm on a subset of 140 plants at three time-points during drought stress, showed that photosynthetic efficiency is less sensitive to drought stress than leaf growth. The candidate genes identified in this study, as well as correlations with field agronomic traits, may provide an insight that helps future understanding of the genetic control of biomass-related traits under both well-watered and drought stress conditions.