The interaction between dissolved organic matter (DOM) and microbial communities plays a crucial yet largely understudied role in biogeochemical cycling within aquatic ecosystems. Recent studies have suggested that key ecosystem functions of lakes—such as carbon sequestration, nutrient dynamics, biodiversity, and thermal regulation—are strongly influenced by the DOM-microbe interactions. A deeper understanding of these interactions is essential for elucidating the chemical and microbial dynamics that can govern ecosystem functioning emerging from their associations. However, related research remains scarce, especially in deep lakes due to logistical constraints. To address this gap, we conducted the first comprehensive analysis of DOM–microbe associations in both the epilimnion and oxygenated hypolimnion of a deep freshwater lake (Lake Biwa, Japan) during the stratification period. We integrated non-targeted ultrahigh-resolution mass spectrometry-based environmental metabolomics with microbiome profiling. To improve the interpretation of DOM-microbe networks, we also developed an integrated compound category classification (IC3) framework for assigning molecular formulae (MFs) of DOM to specific compound categories. Canonical correlation analysis found a significant association between the overall compositions of DOM and microbiome in both the epi- and hypolimnion. Using a compositional data analysis framework, we identified specific MFs and bacterial taxa that co-varied in the hypolimnion during stratification. The hypolimnion exhibited much more complex DOM-microbe networks than the epilimnion, underscoring the stronger coupling between DOM and microbes in the deep. Hypolimnion specialist bacteria were associated with specific MFs, including amino sugar-, carbohydrate-, peptide-, and lipid-like compounds, as well as with other bacteria. These patterns provide environmental metabolomic evidence for substrate preference and potential microbial symbioses. Our study offers the first high-resolution insights into DOM-microbe associations in a deep freshwater lake and establishes a methodological foundation for more efficient and robust analyses of such interactions.