Data obtained using several experimental methods (curing, transconjugation, cloning, and hybridization) indicate that crystal protein genes in many subspecies of BT that are toxic to lepidopterans are located on one or more large plasmids; in some subspecies, the gene may be located on the chromosome. Detailed mapping has shown that in three plasmids (each from a different strain) the genes are surrounded by multiple copies of two repeated DNA elements; the arrangement of these elements is the same in the three plasmids. An analysis of the sequence of one of these repeated DNAs strongly suggests that it contains a transposase. Thus, transfer of crystal protein genes between plasmids and/or between plasmids and the chromosome would be possible either by transposition or by recombinational events mediated by the repeated DNAs. Crystal protein genes have been cloned from several plasmids and were expressed in E. coli and B. subtilis, whereas two genes cloned from chromosomal preparations were not expressed. Some of the factors that regulate expression of a plasmid-borne gene in E. coli and B. subtilis have been identified. Very little is known about the role of sporulation genes in regulating expression of the crystal protein gene in B. subtilis or BT. In BT, expression may also be affected by genes on other plasmids. Three homologous crystal protein genes have been identified and cloned from subsp. kurstaki and thuringiensis; different strains of these subspecies may contain one, two, or three of these genes. It seems probable that additional gene families will be found, since the crystals of different subspecies contain immunologically distinguishable proteins. The DNA sequences of the three homologous genes have been published as has the sequence of the crystal protein gene from subsp. sotto. These four genes have regions of identity (the promoter region) and similarity (the N-terminal approximately 280 amino acids, the C-terminal half of the protein, and the terminator). It is interesting that the divergent portions of the molecules are not in precisely the same positions and that all overlap the toxin-encoding portion of the gene. It would be worthwhile to determine if the differences in the amino acid sequence are related to differences in the toxicity and/or the host range of the cloned genes, and to establish how the complement of genes in a given strain contributes to the overall toxicity of that strain.(ABSTRACT TRUNCATED AT 400 WORDS)