AbstractThe seismogenic process is nonlinear and irreversible so that the response to loading is different from unloading. This difference reflects the damage of a loaded material. Based on this insight, a new parameter-load/unload response ratio (LURR) was proposed to measure quantitatively the proximity to rock failure and earthquake more than ten years ago. In the present paper, we review the fundamental concept of LURR, the validation of LURR with experimental and numerical simulation, the retrospective examination of LURR with new cases in different tectonic settings (California, USA, and Kanto region, Japan), the statistics of earthquake prediction in terms of LURR theory and the random distribution of LURR under Poissons model. Finally we discuss LURR as a parameter to judge the closeness degree to SOC state of the system and the measurement of tidal triggering earthquake.The Load/Unload Response Ratio (LURR) theory was first proposed in 1984 (Yin, 1987). Subsequently, a series of advances were made (Yin and dYin, 1991; Yin, 1993; Yin et al. 1994a, b, 1995; Maruyama, 1995). In this paper, the new results after 1995 are summarized (Yin et al., 1996; Wang et al., 1998a; Zhuang and Yin, 1999).