Background/Objectives: Systematically elucidating the neural mechanisms of bowel and bladder control is crucial for developing closed-loop brain–computer interface (BCI) neurorehabilitation for patients with spinal cord injury (SCI). This study aims to characterize the spatiotemporal Electroencephalography (EEG) dynamics and compensatory cortical reorganization underlying continence motor intentions across healthy aging and SCI cohorts. Methods: 32-channel EEG signals were collected from healthy young adults, healthy older adults, and SCI patients during motor attempt paradigms of suppressed urination and defecation (SUD) and active urination and defecation (UD). Data were analyzed utilizing event-related spectral perturbation (ERSP) and phase-locking value (PLV) functional connectivity across alpha (8–13 Hz), beta1 (13–20 Hz), and beta2 (20–30 Hz) bands. Results: All cohorts exhibited generally consistent broadband event-related desynchronization (ERD) in the sensorimotor cortex (SMC) during visceral motor attempt. Conversely, the frontal cortex (FC) demonstrated marked frequency-band differentiation. During the SUD task, healthy older adults showed significantly greater SMC beta2 ERD than young adults, whereas SCI patients showed a similar but non-significant trend. Frontal beta2 findings were preliminary and should be interpreted with caution due to potential residual artifacts. Network analysis revealed that physiological aging and SCI are associated with large-scale topological reconfigurations, prominently featuring enhanced long-range frontal functional connectivity to compensate for sensory or peripheral deficits. Conclusions: High-level cortical intentions for continence are preserved after SCI, with frontal high-frequency activity possibly reflecting a compensatory process for sensorimotor disruption. These findings substantiate that visceral intentions exhibit discernible characteristics at the cortical level, establishing a spatiotemporal framework for targeted pelvic nerve modulation via brain–machine interfaces.