Hydrothermal liquefaction (HTL) is a promising technique for conversion of wet biomasses containing varying amounts of carbohydrate, protein, lipid, and lignin. In this work, mixtures of these model compounds were subjected to HTL at 335 degrees C. As many as 67 compounds were quantitated in the aqueous phase, including small organic acids, cyclic oxygenates, fatty acids, nitrogenates, and oxygenated aromatics. The concentrations correlated with the ratio of the model compounds. Principal component analysis separated samples on the basis of their quantitative results which could be linked to their biochemical composition. Concentrations of the analytes were modeled with partial least squares regression, and high-quality predictions were made from quality control (QC) samples and to varying degrees from Dried Distillers Grains with Solubles (DDGS), Miscanthus x giganteus, and Chlorella vulgaris. Values for total organic carbon (TOC), total nitrogen (TN), and pH were also predicted from QC samples, DDGS, M. x giganteus, and C. vulgaris. Carbohydrate and lipid contents mainly influenced TOC values and could be used for minimizing loss of organics, for techno-economic analysis, and for assessing potential for anaerobic digestion and thermal gasification. Pyrazines were modeled using linear, exponential, and second-degree polynomial fits, depending on whether carbohydrate or protein was the limiting biochemical component, which could be a way of controlling nitrogen and carbon displacement to the aqueous phase. This work shows that TOC, TN, pH, and concentrations of single compounds in the aqueous phase from HTL can, in many cases, be predicted from HTL of mixtures of biomass constituents.