ABSRACT: The strength of a rock material- is determined in the laboratory on representative standard samples. In the case of a closely jointed rock mass it is not possible to obtain a sample with suitable dimensions to represent the whole rock mass. Since 1980, the empirical Hoek-Brown failure criterion (Hoek and Brown, 1980} began to be used to overcome the difficulties in laboratory shear strength determination of jointed rock masses,. The empirical failure criterion was used in conjunction with the Geomechanics Classification System-RMR (Bieniawski, 1989) until 1994. However, due to the limitations in the RMR classification scheme particularly for very poor quality rock masses and for unrealistic rating adjustments for discontinuity orientation in slopes, the feature criterion has been modified over the years.. Recently,, the originators of the criterion introduced a new index., called Geological Strength index (GSI), in to the criterion. The GSI is based upon the visual impression on the rock mass structure and consists of twenty codes to identify each rock mass category and to estimate the GSI value (Hoek and Brown, 1997). Because rock mass classification requires time consuming procedures and has some limitations, the existing GSI system seems a more practical parameter for the determination of the strength of jointed rock masses from field observations. However,, the system is lack of measurable and more representative parameters, and related interval limits or ratings for describing the structure and surface conditions of discontinuities. This situation results in subjective assessments on the determination of the GSI value: In other words,, it is possible to estimate different GSI values for the same rock mass by different persons. The other importani problem of the criterion is the use of undisturbed and disturbed rock mass categories for determining the parameters in the criterion, for which clear guide Une s are lacking.. It is also noted that the data supporting of the revisions have not been published. These uncertainties make it difficult to judge their validity and performance.. In order to provide a more quantitative basis for evaluating GSI values, the authors of this paper suggested some modifications by introducing easily measurable rock mass parameters with ratings and/or intervals. For the purpose, two terms, namely Structure Rating (SR) based on volumetric joint count and- Surface Condition Rating (SCR) from, the input parameters are introduced in to the GSI system, and the modified GSI chart is established. In the latest version of the failure criterion (Hoek and Brown, 1997; Hoek et al., 1998) average undisturbed in-situ conditions are considered to estimate the GSI without application of any adjustment due to any disturbance effect, such as Masting.. Method of excavation,,, major planes of weakness or change in stress are considered, as local features influencing the rock mass at a particular .location. Therefore,, the influence of such factors should be compensated, and necessary adjustments should be taken in to consideration. In this study, a method was proposed to assess the influence of disturbance on rock constants due to method of excavation as discussed in detail by Sönmez, and Ulusay (1999`),, The modifications and the method suggested have been applied to well studied five slope instabilities from Turkey to check the validity and performance of the modifications and the methodology of parameter estimation, Four cases were selected from the slopes excavated in heavily jointed rock masses and one from spoil piles in a strip coal mine. The application of the suggested modifications and the method examined by the hack analysis of the failures indicated that the use of GSI value determined from the suggested modified chart and consideration of disturbance effect confirmed the limit equilibrium condition for the investigated failed slopes. The back analysis results from a spoil pile instability revealed that spoil pile materials consisting ofblocky and angular rock pieces with small amount of fines could be categorized as a disintegrated rock mass in the GSI system and the criterion seemed to be applied to spoil materials,. However, future applications of the suggested modifications on to failure case studies both from surface and underground excavations may provide a better tool for more precise guidelines and to check the performance of the equations of the criterion.