Background: Postural instability is a restrictive feature in Parkinson’s disease (PD), usually assessed by clinical or laboratory tests. However, the exact quantification of postural stability, using stability theorems that take into account the human dynamics, is still lacking. We investigated the feasibility of control theory, Nyquist stability criterion (Gain Margin – GM –, and Phase Margin – PM –), in discriminating postural instability in PD; as well as the effects of a balance-training program. Methods: Center-of-pressure (COP) data of 40 PD patients before and after a 4-week balance-training program, and 20 healthy control subjects (HCs) (Study1); as well as COP data of 20 other PD patients at four time points during a 6-week balance-training program (Study2), collected in two earlier studies, were used. COP was recorded in four tasks, two on rigid surface and two on foam, eyes-open and closed. A postural control model (an inverted-pendulum with PID controller and time delay) was fitted to the COP data, to subject-specifically identify the model parameters; thereby calculating |GM| and PM for each subject in each task.Results: Patients had smaller margin of stability (|GM|, PM) compared to HCs. Particularly, patients, unlike HCs, showed drastic drop in PM on foam. Clinical outcomes and margin of stability improved in patients after balance training. |GM| improved early at week 4, followed by a plateau in the rest of the training. In contrast, PM improved late (week 6), in a relatively continuous-progression form. Conclusions: Using fundamental stability theorems is a promising technique for standardized quantification of postural stability in various tasks.