Introduction Probabilistic analysis is regarded as a relatively new technique to assess slope stability. In con- ventional limit equilibrium analysis, the deterministic safety factors (Fs) could not reflect actual insta- bility level, because conventional analysis could not establish the relationship between Fs and slope instability probability [1-3]. Today, probabilistic analysis is more and more often applied, as it can help define the risk level of a landslide [4-10]. The main point of this analysis is to determine the Fs probability distribution based on the prob- ability distributions of different influencing factors, among them soil strength, underground water level, seismic coefficients, etc.[3]. Of these, soil strength probability distributions can be achieved by field and laboratory study [10]. Seismic slope stability analysis can be carried out based on either pseudostatic analysis or dynamic analysis [11]. However, each will provide different results. In contrast to pseudostatic analysis, in which only one acceleration value is considered, dynamic analysis takes more acceleration informa- tion into account. Therefore, any direct comparison between the results of the two methods is false. For example, in a dynamic analysis, an earthquake with a relatively small peak ground acceleration can cause greater catastrophic damage than an earthquake with greater acceleration value. In this work, the probability distribution of a seismic coefficient based on the acceleration dis- tribution of seismograms was presented to analyze its influence on slope stability results. The slope stability assessment was carried out under seismic conditions in the cableway station area of a slope located in the northern part of Aibga Ridge. First, a conventional analysis was carried out by means of the pseudostatic method, which led to a deterministic safety factor. Second, a proba- bilistic analysis was carried out in consideration of seismic acceleration distribution. Both steps were completed by applying the Morgenstern-Price method [12], which satisfies both moment and force equi- librium and is the most mathematically strict. The Morgenstern-Price method is the most widely applied in Russian code [13]. Soil Mechanics and Foundation Engineering, Vol. 56, No. 2, May, 2019 (Russian Original No. 2, March-April, 2019) This study sought to assess the probabilistic stability of slope analysis under seismic con- ditions in the middle of the Mzymta River Valley in Russia's Greater Sochi. In proba- bilistic analysis, almost all input parameters can be assigned a probability distribution; however, the seismic coefficient probability distribution has been difficult to define in pre- vious studies. This study aims to assign a probability distribution to the seismic coeffi- cient based on seismogram acceleration distribution. The study showed that seismogram- based probabilistic analysis is a feasible way to assess slope stability. THE ACCELEROGRAM-BASED PROBABILISTIC ANALYSIS OF SLOPE STABILITY K. Kang, 1* O. V. Zerkal, 1 I. K. Fomenko, 2 O. V. Pavlenko 3 1 Geological Faculty, Lomonosov Moscow State University, Moscow, Russia; 2 Faculty of Hydrogeology, Russian State Geological Prospecting University named after Sergo Ordzhonikidze, Moscow, Russia, 3 Schmidt Institute of Physics of the Earth, Russian Academy of Sciences, Moscow, Russia. * Corresponding author E-mail: kevinkang8@mail.ru. UDC 624.131.535 SOIL MECHANICS Translated from Osnovaniya, Fundamenty i Mekhanika Gruntov, No. 2, pp. 2-8, March-April, 2019. 0038-0741/19/5602-0071  2019 Springer Science+Business Media, LLC 71 DOI 10.1007/s11204-019-09572-z