This paper reports on an intensive study into releases of polychlorinated dibenzo-p-dioxins (PCDD), polychlorinated furans (PCDF) and polychlorinated biphenyls (PCB) from a diesel engine and the analysis of PCDD/F and PCB in crankcase lubricating oil. Experimental conditions were set and carefully controlled in order to maximize the possible impact of, and our ability to measure the effect of, changes in the levels of chlorine in the lubricant. Emissions to air were measured using modified EPA methods following the principles of the European EN 1948 standards. A series of 40 experimental runs were completed using three reference lubricants formulated to have three levels of chlorine present as a residual component (at levels of 12, 131 and 259 mg kg(-1) or ppm). The engine was run with and without the diesel oxidation catalyst. All lubricants were realistic oils and the use of unrealistic additives or doping of components - particularly chlorine - in the fuel and lubricant was carefully avoided. Analysis of fuel and lubricant (before and after testing) samples required strenuous attention to achieve acceptable recoveries and showed non-detectable levels of PCB and PCDD/F at a detection limit of around 1.5 ng I-TEQ kg(-1) (ppt), indistinguishable from the laboratory blank. The testing demonstrated the need for extreme care to be taken in developing measurement methods that are sufficiently sensitive for measuring chlorine content of fluids and PCDD/F in oils, the latter being particularly challenging. Mean emissions of PCDD/F with the diesel oxidation catalyst in place were 23 pg I-TEQ l(-1) of fuel and with the diesel oxidation catalyst removed 97 pg I-TEQ l(-1) of fuel. The results of this testing showed that the emissions of PCDD/F were greatly reduced by the presence of a diesel oxidation catalyst in the exhaust, a finding that has not been explicitly tested in previous work. They also show that emissions from the engine were not controlled by the level of chlorine in the lubricant and that emissions did not change in response to a much greater step change in the total chlorine entering the combustion chamber due to a change in the level of chlorine in the fuel. Emissions when the engine was configured with a diesel oxidation catalyst showed a consistent pattern that appears to be unique in the experience of the authors.