E.J. Lowe has a unique view on dispositions which, so far, has not attracted a great deal of attention. I do not advocate Lowe’s picture, but I think certain aspects of it are correct. In particular, I agree that in order to understand dispositionality, we must think about the nature of universals and the second-order relations in which they stand. Lowe’s picture also suggests that disposition manifestation and property instantiation are much the same thing, and again this is an idea that I find attractive. My overall aim in this paper is to assess some of the benefits and costs of Lowe’s position, and then to sketch my own position which keeps the benefits but avoids some of the costs.