The magnitude of initial geostress is one of the key parameters for designing underground structures and can directly influence the mechanical behavior of engineered rock masses. Although attention has been paid to the engineering problems related to high geostresses since 1980s

no clear definition is ever made thus far. In China

several rating schemes have been proposed for defining initial geostresses

but the rockmass behaviors inferred from such schemes differ more or less from the real ones. For example

no obvious high geostress characteristics were ever witnessed in some predicted high-geostress engineering sites (e.g. Guandi underground powerhouse)

whereas heavy rockbursts could happen in the sites supposed to have low geostress (say

Ertan underground powerhouse site). To resolve this issue

the rating schemes for initial geostress commonly practiced in China is briefly introduced first; then the major factors affecting the geostress rating are discussed; finally high geostress is classified into two categories

i.e. the initial high geostress and the induced high geostress. The induced high geostress is a combined result of the high secondary stress concentration due to cavern groups and the dynamic disturbance due to blasting excavations. The criterion defining high geostress is clearly specified

which is the threshold of geostress that can induce failure of the embedded structure or the rockmass

and a qualitative criterion is also summarized based on the previous research results. The ratio of strength to stress is redefined as the uniaxial compressive strength of dry intact rock to the measured maximum principal stress

and a new initial geostress rating scheme (quantitative criterion) is suggested. The suggested scheme is validated using the monitoring data of 25 engineering cases

showing that the accuracy of the suggested scheme is much better than the commonly used rating schemes in China. The suggested scheme is similar to that proposed by CD Martin et al. in 1999.