Some that have shown mild evidence in support of it are typically non-reproducible.
Last I heard there was one bit of weak evidence for it in the perception of trivially different shades of, I believe it was, blue. Then again, that was perhaps 5 years ago, the evidence was new, and perhaps that's been falsified since then.
I find it a bit more than a coincidence that Whorf's analysis was based on not understanding Navajo, was it?
On the other hand, there are large swaths of academia that crucially depend upon it for rationalization. As one believer, put it, though, "That assumes that the goal of experiments is to falsify and not substantiate what you know to be true." Spoken like a true believer. He has a point--the social sciences and humanities don't need to build bridges.
Then again, if engineering had the same track record as education, we'd still be dealing with a 30% bridge failure rate after 60 years of research involving many, many billions of dollars, and be calling it a success.