The use of sub-keV primary ion beams for SIMS depth profiling is growing rapidly, especially in the semiconductor area. The first challenge in this new branch of the technique was to invent equipment capable of obtaining high quality data with sufficient rapidity to be economically viable. After a brief historical introduction, these radical developments across instrument type are reviewed. The current challenge is to obtain a good enough understanding of the experimental process to obtain accurate, interpretable, data. The elementary requirements for this are described with emphasis on data density. Finally, the physical limitations such as atomic mixing and transient effects at the surface and interfaces are discussed. Extrapolation of profile shape as a function of beam energy, combined with the use of capped samples, is discussed as a potential part of the solution. (C) 2002 Elsevier Science B.V. All rights reserved.