Standing balance experiment and the measured data-set are fundamental for identifying postural feedback controllers. As the generalized feedback controllers can only be identified from long duration balance data (under random external perturbations), a standing balance experiment is conducted and the long duration motion data was recorded. The data-set includes the perturbation reaction data from eight subjects. Each subject performed four experiment trials, including two quiet standing and two perturbed trials. Each trial lasted five minutes. A total of 80 minutes quiet standing and 80 minutes perturbed standing data are included in this data-set. Recorded information including three dimensional trajectories of thirty-two markers (27 on subjects' trunk and legs and 5 on the treadmill frame), six dimensional ground reaction forces, and nine Electromyography signals (EMGs, on subjects' right leg). In addition, joint angles and torques were calculated using a human body model and inverse dynamics. Basic statistical analysis of the data is also included. Measured raw data for each subject in each experimental trial includes three files: Mocapxxxx.txt: contains motion capture marker data, ground reaction force, and 76 analog channels. Data was recorded at 100 Hz sampling rate. Mocapxxxx_Motion Analysis_analog.txt: contains 76 high sampling rate (1000Hz) analog channels' data. Analog data is consisted of the analog singal from the froce sensor on the treadmill, EMG signals in the Delsys EMG sensors, and 3 axises acceeleration signals of the Delsys EMG sensors. Recordxxxx.txt: contains the sway motion data of treadmill and the three-axis acceleration data of two Xsens MTi-10 series sensors. Measured raw data also includes two files of the unloaded trial, which is used for the inertia compensation. Mocap0000.txt: contains motion capture marker data (5 markers on the treadmill frame) and ground reaction forces. Record0000.txt: contains the treadmill sway motion data and the acceleration data (three-axis) of two Xsens MTi-10 series sensors. Processed data of each subject in each experimental trial contains four files: Mocapxxxx.txt: contains the gap filled motion capture marker data and the inertia compensated ground reaction force data. Motionxxxx.txt: contains the calculated the trajectories of three joints' (hip, knee, and ankle) angles, angular velocities, moments, and joint contact forces. Data_infoxxxx.txt: contains the quality of recorded raw marker data (percentage and biggest duration of missing marker data), and the percentage of removed inertia artifacts in ground reaction forces MotionAnalysis.fig: shows the mean and standard deviation of three joints' trajectories in four experimental trials. There are two more plots in the processed data folder which shows the joint motion/moment and the raw/compensated ground reaction forces of one example experimental trial (subject 07 trial 03). The processed data was generated using the code in the 'Processing_Code' folder. The code was wrote using Matlab and the main function is "Data_Processing_Main.m" More details of the standing balance experiment can be found in the document 'Standing_Balance_Experiment_with_Long_Duration_Random_Pulses_Perturbation.pdf'