Determining an optimum drug concentration to modulate drug release from a flexible and dynamic system, such as hydrogels requires an extensive amount of experiments consuming time and money. Here, we conduct molecular dynamics (MD) simulations to predict an optimum drug concentration to load on low-methoxyl pectin hydrogels, which can maintain the structural integrity and achieve controlled drug release. Systems of lowmethoxyl pectin hydrogel modeled as poly-galacturonic acid (PGAL) oligomers cross-linked with Ca2+ and procaine at different concentrations (0, 6, 30, 60, 90, 180 mg procaine/g hydrogel) are investigated. First, a series of 200 ns long all-atom MD simulations in explicit water with replicas are conducted reaching a total simulation time of 3.6 mu s. We then assess our predictions with procaine release experiments from low-methoxyl pectin hydrogels, and calculation of storage G' and loss G '' moduli of hydrogel samples. Local and global motions of cross-linked PGAL oligomers from simulation trajectories are analyzed in detail to predict an optimum procaine concentration of 30 mg/g hydrogel. At this loading amount, polymer fluctuations are stabilized and initial monomer neighbors are maintained with a low average distance deviation of 1.5 angstrom so as to yield an intact hydrogel network. Results from MD simulations and experiments strongly agree. Pectin hydrogels loaded with 30 mg procaine/g have a low hydrogel degradation rate 0.001 g/min and a controlled in vitro drug release when compared to other samples, releasing all 30 mg of loaded procaine from 670 mg hydrogel in 24 h. Results indicate that drug release from the flexible host is mainly controlled by local and global motions of the polymer chains and non-bonded interactions between the system components. The computational approach taken here can be highly useful for similar polymer-based drug delivery systems.