The numerous variable valve actuation mechanisms for poppet valves need to be classified, if sensible comparisons are to be made, and one possible taxonomy is presented here. Not all the mechanisms proposed have been tested, but where they have it is usually with gasoline engines. It is well established that controlling the valve events can raise and flatten the torque curve. However, it is difficult to quantify and compare the gains in torque and consequential reduction in fuel consumption, as the results depend very much on the starting point. This is also the case when variable valve actuation is used to reduce engine emissions. Fortunately it is quite easy to realise suitable variable valve timing systems for controlling the valve overlap, and the point of inlet valve closure.The other main application to gasoline engines, is in obtaining load control without throttling. The thermodynamic background is discussed here along with some experimental results; the section ends by assessing the suitability of some of the mechanisms for this purpose.Finally, the application of variable valve timing to diesel engines is discussed. It is argued that the control of valve overlap can flatten the torque curve of some diesel engines, and that the control of the inlet-valve-closing angle should lead to improved starting. However, the most promising application on diesel engines, is the control of the valve overlap on highly turbocharged engines.Since an earlier review of variable valve timing (l)*, there has been a proliferation of both mechanisms and and papers on variable valve actuation (VVA). However, efforts are still almost exclusively directed towards spark ignition engines, in order to increase the torque curve, and to reduce the part load throttling losses. None the less, attention will be drawn here to other uses of variable valve actuation, and some possible diesel engine applications. Work is still also being undertaken on the potential for reducing emissions by using VVA.It is difficult to make a paper such as this both compact and comprehensive; inevitably there are omissions. The next section categorises the mechanisms and illustrates each main type by a typical example. Where possible, purely theoretical studies are identified as such. The various reasons for applying variable valve timing are discussed in the subsequent sections, and the different references are segregated, as to whether they contain experimental results or theoretical/computational predictions.In the section on gasoline engines, particular attention is paid to the reduction of the throttling losses at part load, and it is argued that few of the proposed mechanisms are suited to this task. There is also a discussion of the role of inlet valve closure and valve overlap on gasoline engine performance, and the mechanisms that might achieve these variations.In the following discussion of variable valve actuation systems mechanisms that have been produced and tested are identified as such.