Lean burn and EGR are two commonly used technologies for improving fuel consumption and controlling emissions. Each has advantages and disadvantages when applied to an engine and, qualitatively at least, these effects are well known. To meet future emissions and fuel consumption constraints, most production engines are likely to use one or the other. This paper describes testing undertaken to quantify the opportunities that each strategy offers on a production engine.The engine used in this test program was a modern, four valve per cylinder l-4, with high charge activity features including one high swirl port per pair and one straight port with a deactivation plate. This engine is designed to operate with mixtures as lean as 24:1 during normal operation. It was tested at a number of part load operating conditions typical of an emissions drive-cycle over a range of air-fuel ratios. The engine was then fitted with an EGR system and re-tested at the same speed-load conditions, operating with a stoichiometric strategy and varying quantities of EGR.Results are presented which directly compare the tradeoff between emissions, fuel economy and combustion stability for the two systems. They show that this engine could tolerate at least 25% EGR at most operating conditions, achieving a 90-95% reduction in NOx compared with stoichiometric operation whilst maintaining reasonable control of HC and good combustion stability. Lean burn on the other hand achieved more moderate NOx control and had poorer combustion stability and HC emissions. However, using lean running, the best possible specific fuel consumption was between 3 and 10% lower than what could be realised using EGR.