How is it that a speaker S can at once make it obvious to an audience A that she intends to communicate some proposition p, and yet at the same time retain plausible deniability with respect to this intention? The answer, I suggest, is that S can bring it about that A has a very high justified credence that ‘S intended p‘ without putting A in a position to know that `S intended p‘. In order to achieve this, I argue, the speaker has to exploit a sense in which communication can be lottery-like. Two objections to this proposal are considered: 1. It results in an unacceptable testimonial skepticism, 2. it conflicts with the standard picture of understanding and communicative success. I argue that the resultant skepticism is limited and plausible. And I suggest that the standard picture of understanding and communicative success is wrong.