Measured values of velocity and associated turbulence properties have been obtained in a motored single-cylinder Diesel engine and in a plexiglass simulation engine. The quality of the signals obtained from thereal engine and the corresponding useful data rate are quantified and discussed. As a consequence, it is shown that measurements in real engines are likely to relate to crank-angle intervals of the order of 10 degrees. The implications of this conclusion are quantified by analysis and by measurements in the plexiglass engine.The results, from the motored Diesel engine, also show that values of velocity cannot be measured throughout the cylinder cavity. It is argued that best advantage can be obtained from laser-Doppler anemometry by conducting a small number of measurements in a real engine; demonstrating that these correspond closely to corresponding measurements in a plexiglass engine with the same geometrical arrangement; and using the plexiglass engine to determine the influence of important parameters such as piston-crown shape, head shape, valve arrangements etc.