AAs a prominent component of turbidite deposition systems, turbidite lobe internal architecture characterization has proven essential due to its complicated sedimentary hierarchy and evident heterogeneity. This article demonstrates an integrated methodology for doing multiple point stochastic (MPS) simulation of deep-water turbidite single lobe architecture. Based on logging data, high-frequency seismic data, thorough architectural features analysis and 3D training image, MPS based geological modelling of Miocene turbidite lobe reservoir in Lower Congo Basin are carried out. This effort has two objectives: (1) to expand the geological knowledge base of deep-water turbidite lobes with morphology parameters and (2) to develop a process of turbidite geo-modelling that could characterize the architectural hierarchy of a single lobe with limited hard data. As a first step, we analyze and characterize properties of single lobe elements characteristics and the manner of sedimentary dispersion using 145-meter-long cores, well logging, and seismic analysis. Second, shallow seismic-based turbidites lobes pick-up and measurements to collect quantitative characteristics of turbidite lobes morphology has been conducted and will be used as geo-modelling guidance. Thirdly, a 3D lobe complex training image with single lobe architecture elements superposition is derived by seismic geo-body caving (using threshold truncation) and enhanced based on sedimentary distribution mode. MPS simulation incorporating well data, morphological parameters, training image and seismic inversion constraint is then performed, resulting in an architecture model that could describe single lobes is obtained. The simulation results generally followed the lobe architecture elements morphology and superposition. The coincidence between the MPS simulated turbidites lobe complex architecture model and the posterior well that could reach up to 86%. The article gives a methodology for a case study that proved the implementation of single turbidite lobe architectural characterization using multiple point stochastics, and the recommended process could be applied to other fields.