Direct injection compression ignited (CI) engines are today's most efficient engine technology, granting efficiencies exceeding 40% for their optimal operation point. In addition, a strong technological development has allowed the CI engine to overcome its traditional weak points: both its pollutant emissions and the gap in specific power regarding its competitor, i.e. the spark ignited (SI) engine, have been noticeably reduced. Particularly, the increase in specific power has led to the downsizing as an effective method to improve vehicle efficiency. Despite the reduction in total displacement, the cylinder displacement of current CI engines is still around 0.5 liters. For some applications (urban light duty vehicles, Range Extenders…) it may be interesting to reduce the engine displacement to address power targets around 20kW with high efficiencies. This paper assesses the thermo- and fluid-dynamic limitations which make challenging extending the application of automotive CI engines to the low power region: Firstly, space limitations for injection and combustion processes. In second place the increase of surface-to-volume ratio which gives rise to higher heat losses. Finally, limits related to the air management, most notably with turbocharging, due to the reduction in turbocharger efficiency with decreasing size.The combination of scaling laws with parametric studies carried out with CFD tools allows quantifying the impact of the previous issues on the engine efficiency. The paper also addresses which would be the minimum size of a CI engine for automotive applications (specific power around 40KW/liter and 40% efficiency) with state of the art technology, which according to the authors, is bounded at 150cm3 for cylinder displacement and 450-500cm3 for total displacement due to injection and turbocharging limitations, respectively. Finally, the authors propose a design of such a minimum size engine, which is currently under construction.