Despite systemic therapy and cystectomy, bladder cancer is characterized by a high recurrence rate. Serum glycomics represents a promising source of prognostic markers for monitoring patients. Our approach, which we refer to as "glycan node analysis", constitutes the first example of molecularly "bottom-up" glycomics. It is based on a global glycan methylation analysis procedure that is applied to whole blood plasma/serum. The approach detects and quantifies partially methylated alditol acetates arising from unique glycan features such as α2-6 sialylation, β1-4 branching, and core fucosylation that have been pooled together from across all intact glycans within a sample into a single GC-MS chromatographic peak. We applied this method to 122 plasma samples from former and current bladder cancer patients (n = 72 former cancer patients with currently no evidence of disease (NED); n = 38 non-muscle invasive bladder cancer (NMIBC) patients; and n = 12 muscle invasive bladder cancer (MIBC) patients) along with plasma from 30 certifiably healthy living kidney donors. Markers for α2-6 sialylation, β1-4 branching, β1-6 branching, and outer-arm fucosylation were able to separate current and former (NED) cases from certifiably healthy controls (ROC curve c-statistics ~ 0.80); but NED, NMIBC, and MIBC were not distinguished from one another. Based on the unexpectedly high levels of these glycan nodes in the NED patients, we hypothesized that recurrence of this disease could be predicted by some of the elevated glycan features. Indeed, α2-6 sialylation and β1-6 branching were able to predict recurrence from the NED state using a Cox proportional hazards regression model adjusted for age, gender, and time from cancer. The levels of these two glycan features were correlated to C-reactive protein concentration, an inflammation marker and known prognostic indicator for bladder cancer, further strengthening the link between inflammation and abnormal plasma protein glycosylation.