The 401(k) regulations specifically preclude you from permissively aggregating 401(k) plans with inconsistent testing methods (safe harbor and non-safe harbor) for testing purposes. Perhaps this is obvious, but does this provision preclude you from permissively aggregating the plans and test ing them on an aggregated basis using the the ADP test - In other words, ignoring the safe harbor provisions??

I don't think there is anything to prevent you from aggregating them (since the rules say you can aggregate plans), but you do lose the safe harbor, and I assume top-heavy free ride. plus the other plan had better be using current year testing since a safe harbor plan is deemed to use current year testing.

That is an excellent point. I apologize in advance for asking you this, but where does it say that safe harbor plans are deemed to use current year testing. I know that is the correct result, and will look for (and post!!) the cite if you do not know it off the top of your head.

its buried in 1.401(k)-2©(1)(i) - also I think you would find it in the regs under discontinuing the safe harbor, because of of the stipulations is that you test use the current year method.

and also, since a safe harbor contribution is a 'QNEC' or "QMAC", I think the ultimate logic would be that you are providing something based on 'current year' comp, so logically use current year testing.