Patient random irregular breathing undermines the fundamental assumption for 4DCT and radiotherapy based on it, either ITV-based or gated treatments. During patient’s 4DCT simulation, Respiratory Gating Signal Waveform (RGSW) is usually collected over a duration much longer than one breathing cycle. Therefore, RGSW may contain valuable information on irregularities in the patient breathing cycles and help to improve imaging and delivery in 4D radiation therapy. In this study, we developed a RGSW based tool for quantitatively accessing the randomness and irregularity level in patient respiratory motion. Totally 242 clinical RGSWs obtained with Varian Real-time Position Management (RPM) system were included in this study. For each RGSW, firstly we constructed its unwrapped phase φ(t) using Hilbert transform. Secondly, we computed the phase differences, i.e., φ(t+τ)-φ(t), between any two points along the waveform with a fixed time interval τ. Thirdly, we estimated the standard deviation σ of the phase differences. Finally, we calculated volatility (i.e., σ ⁄√τ) of the RGSW's phase in the given time interval τ. Sixteen intervals (i.e., τ=1s, 2s, 3s, …, 16s) were selected for volatility’s computation to examine if it was stable over different time intervals. The relative standard deviation of the acquired volatility values at all 16 time-intervals was computed for each RGSW. Fittings were performed on the histograms formed by observed volatility over a fixed time interval (e.g., 4s) and the relative standard deviation in its measured values with all 16 intervals from all patient RPM RGSWs. A stochastic random walk phase model was utilized to interpret the results. The unwrapped phase of an ideal periodic RGSW always has zero volatility. For our clinical RGSWs, the volatility is non-zero but reasonably stable over the selected time intervals. The observed volatility at 4s interval from the 242 RGSWs (∈ [0.07, 3.7] (1⁄√s), median=0.78 (1⁄√s)) could be approximated by a Gamma distribution Γ(3.0, 0.3). The population histogram of the relative standard deviation in volatility for the investigated time intervals (∈[2%, 35%], median=11.5%) could be fitted with a Gamma distribution Γ(3.5, 0.04). The stochastic random walk phase model, which yields a fixed volatility value over any given time interval, might serve as a useful tool in explaining our observations on phase volatility in patient's RGSW. We proposed a new application of patient’s RGSW for measuring the randomness and irregularity in patient respiratory motion. The volatility in RPM RGSW phase has a potential correlation to the eligibility and treatment response of a given patient to a given clinical procedure. Further studies are needed to explore its potential clinical role in patient 4D image acquisition and patient radiotherapy delivery.