Deposition of small metal-oxide clusters on the Zr-based nodes of a metal-organic framework has been demonstrated to provide access to a variety of single-site catalysts. Well-defined catalytic active sites are amenable to detailed computational studies of potential catalytic pathways, and they invite screening a wide range of metals to assess their expected activity. Here we report the application of density functional theory to a variety of transition metals (in particular Ti-IV, V-II, V-IV, Cr-II, COIII, Mn-II, Mn-IV, Fe-II, Fe-III, Ni-II, Co-II, Co-III, Cu-II, Cu-III, Pd-II, Mo-II, and W-II) supported on NU-1000 inorganometallic nodes to evaluate their activity for ethylene dimerization. We found that the rate-determining step varies between different catalysts, which illustrates the importance of considering more than a single step when comparing catalytic cycles across a variety of metals. Our calculations are consistent with the known good activity of supported Ni-II for ethylene dimerization, and they predict that Cr-II and Pd-II are also potentially useful catalysts for this process. We also screen modifications to the organic linker of Ni-II-NU-1000 by considering the addition of Me, iPr, tBu, CF3 and NH2 groups to study the influence of sterically demanding and, the case of CF3 and NH2, respectively, electron-donating and withdrawing, substituents on the activity for ethylene dimerization; we predict no improvements in activity or selectivity (for 1-butene) with such substitutions. (C) 2017 Elsevier Inc. All rights reserved.