Depends on the size and reasons for the rewrite. I am Director of Eng. at a 85 people startup (funny how that sounds), and last April we released "version 2" o

Depends on the size and reasons for the rewrite. I am Director of Eng. at a 85 people startup (funny how that sounds), and last April we released "version 2" of our platform/website [1]. We basically threw away a huge chunk of the code that was developed by the technical founder, myself and another dev, and replaced it with a fully scalable architecture. When the company I work for was being started, a technical advisor (who had sold his company months before) told us not to worry a lot about the first codebase, because we were going to throw that away after some time. I am deffinitely sure even what we have now, will get obsolete in a couple of years (two at most) For this reason, I think having a good planned rewrite (even if it is underestimated) can be good. I think the important part is understanding why the rewrite is performed, and be sure that it is done for the right reasons. [1] http://nerds.kueski.com/the-path-to-kueski-2-0/