Human-structure interaction is a relatively new topic that is not fully understood. There have been several human whole-body models from the research in body biomechanics and structural dynamics, which have been used in the study of human-structure interaction. It is not clear which body model is the most appropriate one. An interactive human body model was derived from a human-structure interaction model where a continuous standing human body was placed on a single degree-of-freedom (SDOF) structure. However, the dynamic parameters of the human body model cannot be determined accurately. In this thesis, a series of human-structure interaction experiments are conducted, which also leads to the identification of the dynamic parameters of the interactive body model and the assessment of the commonly used human body models. Two groups of 18 and 38 individual subjects participated in human-structure interaction experiment on a SDOF test rig with two different configurations. Two sweeping harmonic forces (6.6 and 13.2 N) were applied to the bare and occupied rigs. The repeatability of the tests was checked and confirmed. These experiments showed clearly two resonance frequencies of the human-structure system. It was also demonstrated that the dynamic parameters of the standing human body were independent of the test rig setup and of the subjectsâ gender. On the other hand, the vibration magnitude and the body masses significantly influenced the natural frequencies but not the damping ratios of the standing subjects. The fundamental natural frequency and damping ratio of the standing human body were about 6.6 Hz and 22% respectively. The identified dynamic parameters of the standing body can then be used to predict the responses of an occupied structure and the human body. Another group of 74 subjects were tested twice, with and without wearing shoes, which examined the effect of footwear on the dynamic parameters of the standing human body and on the dynamic response of the occupied rig. Only one sweeping harmonic force (13.2 N) was applied to the test rig. This study demonstrated that footwear significantly affected the dynamic parameters of the standing human body. The natural frequency and damping ratio of the standing body with bare feet are higher than those with footwear. When the two genders have the same body mass index (BMI), the maximum responses of the occupied rig are almost identical. When they have the same weight, the response of the rig occupied by the males was higher at the first resonance peak. The accelerations throughout the heights of two subjects were measured, which allowed a comparison between the predicted human whole-body acceleration and the measurements at different positions of the standing human bodies. The predicted frequency response functions (FRFs) had the same pattern as the measured ones and were larger than the measured responses at the head, neck and shoulders. The effects of the mass ratio of a crowd to a SDOF structure and the natural frequency of the structure on the human-structure interaction were examined. It was demonstrated that, for a light crowd, such as seen on office floors, the occupied structure would respond less than the bare structure, where the human body acts like a tuned-mass-damper, while the body responses were higher than that of the bare structure. For a larger crowd, such as seen on grandstands, the responses of the occupied structure and the human body were both smaller than that of the bare structure although the body response was larger than that of the occupied structure. A comparison between the human-structure interaction model used in this study and three other models was conducted. The dynamic parameters of the models were identified from the above experiments, in which the natural frequencies of the body for the four models were similar. It showed that the predicted responses of the occupied structure were similar based on the four models. However, there were obvious differences in the predicted body responses. A detailed comparison between the proposed model, Griffinâs models and the available measurements showed that the damping ratios used in Griffinâs models were too high, which prevents the two resonance frequencies from being observed. In addition, the predicted human body response calculated by the proposed model is much higher than that from Griffinâs models.