In this discussion of Alan Kindler's article, I examine a number of principles pertaining to improvisation as a mode of play in psychoanalytic treatment. These include differentiating improvisation from spontaneity, a series of questions germane to tracking improvisation referred to as a relational ethic, and the concept of the positraum. In elaborating my discussion of Kindler's cases, I explicate three versions of improvisational play, improvisation, the playful to and fro within the treatment; mutual inductive identification, the lapsing by patient and analyst into unconsciously scripted versions of their engagement; and, finally, Improvisation, the untangling and illuminating of the dyad's enactments. This third mode sometimes culminates in what I refer to as positraums, which are potentially conviction-altering moments that defy personality organization, thereby providing an experience of something positively unimaginable such as initiating new convictions.