Following landmark legislation passed more than 20 years ago, university-industry relationships have now become central to understanding the changing role of research universities in American Society. The paper analyzes the development of university-industry partnerships during the 1990s. Past studies have used a broad array of measures of ties and a variety of research methodologies, but they have shared a focus on top collaborators or on samples of universities skewed toward the top. However, findings based on top collaborators may not be valid for other universities. Universities involved in mid- to low-levels of collaboration are qualitatively different in many ways from the more extensively studied set of top collaborators, suggesting that characteristics affecting university-industry ties may not be the same for these institutions. The paper shifts the focus away from top collaborators to this sizable and less studied majority. In general, we find that the same characteristics predicting high levels of involvement for the sample as a whole also predict high levels of involvement for the sub-sample of mid- and low-level collaborators. However, We find some particular characteristics of these institutions, such as land grant status, are also associated with stronger ties to industry, and that some characteristics of 25 top institutions do not predict the level of involvement of these lower-level collaborators. We will discuss whether the licensing of new technology is likely ever to become an important source of net revenues for current middle and low-level collaborators. Our findings raise doubts about whether many universities below the top 25 will earn substantial net revenues from licensing, though they do not dispute the potential service value of these ties. The study is based on examination of a wide range of potential influences on university-industry collaboration for institutions that are not currently among the most heavily involved in partnerships. These include status, other institutional characteristics (such as size and control), investment in science and engineering, and characteristics of offices or technology transfer.