Sigma-F, the first sporulation-specific transcription factor of Bacillus subtilis, is regulated by an anti-sigma factor SpoIIAB, which can also act as a protein kinase that phosphorylates the anti-anti-sigma factor SpoIIAA. The time course of phosphorylation reaction is biphasic, a fact that has been interpreted in terms of a mechanism for sequestering SpoIIAB away from sigma (F) and thus allowing activation of sigma (F) when needed. Site-directed mutagenesis of SpoIIAA has allowed us to isolate two mutants that cannot activate sigma (F) and which are therefore Spo(-). The two mutant SpoIIAA proteins, SpoIIAAL61A and SpoIIAAL90A, are phosphorylated with linear kinetics; in addition they are less able to form the stable non-covalent complex that wild-type SpoIIAA makes with SpoIIAB in the presence of ADP. The phosphorylated form of SpoIIAAL90A was hydrolysed by the specific phosphatase SpoIIE at the same rate as wild-type SpoIIAA-P, but the rate of hydrolysis of SpoIIAAL61A-P was much slower. The secondary structure and the global fold of the mutant proteins were unchanged from the wild type. The results are interpreted in terms of a model for the wild type in which SpoIIAB, after phosphorylating SpoIIAA, is released in a form that is tightly bound to ADP and which then makes a ternary complex with an unreacted SpoIIAA. We propose that it is the inability to make this ternary complex that deprives the mutant cells of a means of keeping SpoIIAB from inhibiting sigma (F).