The interaction of Mn2+, substrates and initiators with RNA polymerase have been studied by kinetic and magnetic resonance methods. As determined by electron paramagnetic resonance, Mn2+ binds to RNA polymerase at one tight binding site with a dissociation constant less than 10 muM and at 6 +/- 1 weak binding sites with dissociation constants 100-fold greater. The binding of Mn2+ to RNA polymerase at both types of sites causes an order of magnitude enhancement of the paramagnetic effect of Mn2+ on the longitudinal relaxation rate of water protons, indicating the presence of residual water ligands on the enzyme-bound Mn2+. A kinetic analysis of the Mn2+-activated enzyme with poly(dT) as template indicates the substrate to be MnATP under steady-state conditions in the presence or absence of the initiator ApA. ATP and UTP interact with the tightly bound Mn2+ to form ternary complexes with approximately 50% greater enhancement factors. The dissociation constant of MnATP from the tight Mn2+ site as determined by longitudinal proton relaxation rate (PRR) titration (4.7 muM) is similar to the KM of MnATP in the ApA-initiated RNA polymerase reaction (10 +/- 3 muM) but not in the ATP-initiated reaction (160 +/- 30 muM). Similarly, the dissociation constant of the substrate MnUTP from the tight Mn2+ site (90 muM) is in agreement with the KM of MnUTP (101 +/- 13 muM) when poly[d(A-T)]-poly[d(A-T)] is used as template, indicating the tight Mn2+ site to be the catalytic site for RNA chain elongation. Manganese adenylyl imidodiphosphate (MnAMP-PNP) has been found to be a substrate for RNA polymerase. It has the same affinity as MnATP for the tight site but, unlike the results obtained with MnATP, the enhancement is decreased by 43% in the enzyme Mn-AMP-PNP complex. These results suggest that the enzyme-bound Mn2+ interacts with the leaving pyrophosphate group. The initiators ApA and ApU and the inhibitor rifamycin interact with the enzyme-Mn2+ complex producing small (15-20%) decreases in the enhancement. The dissociation constant of ApA estimated from PRR data (less than or equal to 1.5 muM) agrees with that determined kinetically (1.0 +/- 0.5 muM) as the concentration of ApA required to produce half-maximal change in the KM of MnATP. In the presence of the initiation specific reagents ApA, ApU, or rifamycin, the affinity of the enzyme-Mn complex for ATP or UTP shows little change. However, ATP and UTP no longer increase the enhancement factor of the tightly bound Mn2+ but decrease it by 30-55%, indicating a change in the environment of the Mn2+-substrate complex on the enzyme when the initiation site is either occupied or blocked. Although the role of the six weak Mn2+ binding sites is not clear, the presence of a single tightly bound Mn2+ at the catalytic site for chain elongation which interacts with the substrate reinforces the number of active sites as one per molecule of holoenzyme and provides a paramagnetic reference point for further structural studies.