We have characterized a series of amber mutations in the A gene of bacteriophage Mu encoding the phage transposase. We tested different activities of these mutant proteins either in a sup0 strain or in different sup bacteria. In conjunction with the results described in the accompanying paper by Bétermier et al. (1989) we find that the C-terminus of the protein is not absolutely essential for global transposase function, but is essential for phage growth. Specific binding to Mu ends is defined by a more central domain. Our results also reinforce the previous findings (Bétermier et al., 1987) that more than one protein may be specified by the A gene.