Biological wastewater treatment relies on microorganisms that grow as flocs, biofilms, or granules for efficient separation of biomass from cleaned water. This biofilm structure emerges from the interactions between microbes that produce, and are embedded in, extracellular polymeric substances (EPS). The true composition and structure of the EPS responsible for dense biofilm formation are still obscure. We conducted a bottom-up approach utilizing advanced glycomic techniques to explore the glycan diversity in the EPS from a highly enriched “Candidatus Accumulibacter” granular sludge. Rare novel sugar monomers such as N-Acetylquinovosamine (QuiNAc) and 2-O-Methylrhamnose (2-OMe-Rha) were identified to be present in the EPS of both enrichments. Further, a high diversity in the glycoprotein structures of said EPS was identified by means of lectin based microarrays. We explored the genetic potential of “Ca. Accumulibacter” high quality metagenome assembled genomes (MAGs) to showcase the shortcoming of top-down bioinformatics based approaches at predicting EPS composition and structure, especially when dealing with glycans and glycoconjugates. This work suggests that more bottom-up research is necessary to understand the composition and complex structure of EPS in biofilms since genome based inference cannot directly predict glycan structures and glycoconjugate diversity.