Is there still a B→πK puzzle?

We perform a fit to the 2006 B→πK data and show that, when one includes (reasonable) theoretical input, there is a disagreement with the Standard Model. That is, the B→πK puzzle is still present. In fact, it has gotten worse than in earlier years. Assuming that one new-physics (NP) operator dominates, we show that a good fit is obtained only when one type of NP is added (this includes a modification of the electroweak penguin amplitude). The NP amplitude must be sizeable, with a large weak phase.