Spotting circumbinary planets around non-transiting binaries

[PlutonianEmpire]

Would it be possible, with current instrumentation, to very carefully watch the orbits of a close binary, and look for minute variations in the stars' orbits around each other due to the tug of a large planet?

I was thinking maybe that if we saw the binary suns take ever-so-slightly different paths each time they orbit each other, we might be able to extrapolate from that and find a large planet being the culprit...

Is it possible with current instrumentation?

Or am I being overly optimistic?

[ajtribick]

It's been done, as far back as 1993.

Despite what NASA has put out in its news releases, the Kepler circumbinary planets are not the only known circumbinary planets, nor were they the first. I'm not going to speculate on why they feel the need to distort the relatively short history of exoplanet discovery...