Protein is animportant macro molecular organic compound that body need tomaintain itself and grow. Proteins are linear polymers made of 20kinds of amino acids linked by peptide bonds, which fold intounique 3-dimensional structures. Biological functions are closelyrelated to protein structure dynamics, and different proteinstructures generate accordingly different biological functions inbiochemistry process. For a protein structure, structure dynamicsmodels can produce protein movement trajectories, or decomposecollective motions into a combination of independent harmonicoscillators with different frequencies. In this way, many of thosefunctional motions of active sites can be analyzed in detail andunderstood that occur in numerous biological phenomena, such asligand binding and catalytic activity of enzymes. This studyelucidates the latest advances in protein structure harmonicdynamics. Three important physical models are analyzed, andcomputation methods are discussed for protein dynamics. Detailedformulations are given for harmonic dynamics model, includingatomic normal modes analysis (NMA), elastic network model (ENM)and other dynamics methods. Some applications of structuredynamics methods are presented with respect to biologicalresearch. Highlighted by recent advances in protein structuredynamics, it is clear that theoretical methods based on physicalprinciple and simplified models can lend important supports tofrontier research in biology, and protein structure dynamicsbecomes now an important research direction in protein folding andprotein structure prediction.