An investigation of the mechanism causing in-cylinder soot deposition in a direct injection diesel engine was carried out. First, an analytical study was undertaken to determine which of following possible deposition mechanisms, thermophoresis, Brownian diffusion, turbulent diffusion, inertial impingement, or electrophoresis were responsible for the deposition of the soot on the combustion chamber walls. Based on a series of numerical models comparing each mechanism under conditions typical of diesel engine combustion, thermophoresis was singled out as the most likely cause of in-cylinder soot deposition.Second, an experiment was performed to test the hypothesis that the soot deposition was caused by thermophoresis. An optical probe was designed to fit an access port in the cylinder head of a Cummins NH250 single cylinder test engine. The probe was designed such that a sapphire optical access window was positioned flush with the combustion chamber surface when the probe was installed. The probe incorporated the capability of changing the temperature of the window surface approximately 100øC when the engine was operating at: 1300 rpm, an equivalence ratio of 0.3, and an intake pressure of 27 psi.A thermocouple was placed near the window to measure the temperature. A fiber optic cable carried the combustion radiation signal, transmitted through the window, to a photodiode detector. The photodiode signal was amplified and then recorded by a high speed data acquisition system. In the testing sequence, the engine was first motored at 1300 rpm and then the fuel was turned on. The radiation signal was then recorded for the next 1600 cycles and was observed to attenuate as soot accumulated on the window. The transmittance of the soot layer on the window, , was calculated by assigning the initial peak voltage radiation signal recorded to Io and the signal measured subsequent to this to I[. Using light scattering theory, the transmittance was then applied to the calculation of the mass deposited on the window. The results from the experimental study show that the soot deposition rate was, on average, 46.9% higher when the window was cooled versus uncooled. These results were consistant with the earlier predictions that thermophoresis is a main contributor to the soot deposits on the in-cylinder surfaces.