A method to calculate white smoke during starting was developed using a total balance of fuel injected and fuel burned. An accurate needle lift sensor with an in situ calibration was designed and used to measure cyclic fuel injection. The effects of ambient temperature, fuel type, injection timing and the number of repeated starting attempts were studied with regard to white smoke formation, cyclic fuel injection and fuel burned. It was found that the colder the ambient temperature, the less unburned fuel was emitted to the atmosphere due to the decrease in cyclic fuel injection. The more volatile the fuel, the easier it was to start the engine at low temperatures, and the less white smoke was produced. Earlier timing of fuel injection during starting resulted in an increased likelihood of engine starting and less white smoke formation. The experiments were conducted on a single cylinder diesel engine, fully instrumented, and situated in a cold room where temperatures were varied and controlled using a microprocessor.