Catalyst deactivation by coke deposition is a problem of great and continuous concern in residue hydroprocessing operations. To reach a better understanding of the very complex nature of coke it is desirable to obtain information on the amount and composition of both the insoluble (IS) and soluble (S) fractions of coke. A standard procedure, which could be adapted by all laboratories involved in investigation of spent catalysts, would make the results obtained more meaningful. Accelerated solvent extraction in comparison with Soxhlet extraction was evaluated concerning the efficiency and relevancy (recovery, experiment time and volume of solvent used) for coke extraction. The accelerated solvent extraction reached the same efficency as Soxhlet after 5 min of extraction compared with 6-12 h by using less solvent. Four solvent-defined coke fractions, such as heptane- (HEP-S), toluene-(TOL-S), tetrahydrofuran- (THF-S), and dichloromethane-soluble (DCM-S) coke, were obtained from spent catalysts used in atmospheric residue hydroprocessing. For a start of run catalyst (I h-240 h) the S-coke decreases rapidly from 75% (1 h) to 30% (240 h). The sequence of yields is THF-S (external coke precursors) < TOL-S (asphaltenes) < DCM-S (internal coke precursors). The catalysts from the end of run contained about 30 wt % HEP-S coke (carryover of feed and product). Depending on the location of the catalyst in the hyrdrotreating unit, TOL-S coke amounted 3-10 wt %. THF-S and DCM-S coke made up to 10 wt % of the total carbonaceous deposit. Temperature-programmed oxidation and C-13 nuclear magnetic resonance were used to study the nature of the soluble as well as the insoluble coke fractions.