Terminase, the DNA packaging enzyme of phage lambda, binds to lambda DNA at a site called cosB, and introduces staggered nicks at an adjacent site, cosN, to generate the cohesive ends of virion lambda DNA molecules. Terminase also is involved in separation of the cohesive ends and in binding the prohead, the empty protein shell into which lambda DNA is packaged. Terminase is a DNA-dependent ATPase, and both subunits, gpNu1 and gpA, have ATPase activity. cosB contains a series of gpNu1 binding sites, R3, R2 and R1; between R3 and R2 is a binding site, I1, for integration host factor (IHF), the Escherichia coli DNA bending protein. In this work, a series of mutations in Nu1 have been isolated as suppressors of cosB mutations. One of the Nu1 mutations is identical to the previously described Nu1ms1/ohm1 mutation predicted to cause the change L40F in the 181 amino acid-long gpNu1. Three other Nu1 missense mutations, the Nu1ms2 (L40I), ms3 (Q97K) and ms4 (A92G) mutations, have been isolated; the relative strengths of suppression of cosB mutations by the Nu1ms mutations are: ms1 > ms2 > ms3 > ms4. The Nu1 missense mutations all affect amino acid residues that lie outside of the putative helix-turn-helix DNA binding motif of gpNu1. The Nu1ms1 and Nu1ms2 mutations alter an amino acid residue (L40) that lies directly between two segments of gpNu1 proposed to be involved in ATP binding and hydrolysis; thus these mutations are likely to alter the gpNu1 ATP-binding site. The Nu1ms3 and Nu1ms4 mutations both affect amino acid residues in the central region of gpNu1 that is predicted to form a hydrophilic alpha-helix. To explain how the Nu1ms mutations suppress cosB defects, models involving alterations of the DNA binding and/or catalytic properties of terminase are considered. The results also indicate that terminase occupancy of a single gpNu1 binding site (R3) is necessary and sufficient for the efficient initiation of DNA packaging; the Nu1ms1, ms2 and ms3 mutations permit IHF-independent plaque formation by a phage lacking R2 and R1.