Genomic DNA from type A Clostridium botulinum was digested with restriction endonucleases that cut at rare sites, and the large fragments were separated by pulsed-field gel electrophoresis. Of 15 restriction enzymes tested, MluI, RsrII, SmaI, NruI, KspI, NaeI, and XhoI generated satisfactory digestion patterns of genomic DNA of various C. botulinum strains, enabling the use of the method for genomic fingerprinting. The genomes of four group I (type A) C. botulinum strains examined had similar restriction patterns. However, each strain had unique digestion patterns, reflecting genotypic differences. The genome size of C. botulinum strain 62A was estimated to be 4,039 ؎ 40 kbp from the summation of restriction fragments from MluI, RsrII, and SmaI digestions. Genes encoding proteins involved in the toxinogenicity of C. botulinum, including neurotoxin, hemagglutinin A, and genes for a temperate phage, as well as various transposon Tn916 insertion sites in C. botulinum 62A, were mapped by pulsed-field gel electrophoresis. The genes encoding neurotoxin and hemagglutinin A-1, were located on the same fragment in several cases, indicating their probable physical linkage. The macrorestriction analysis established here should be useful for genetic and epidemiological studies of C. botulinum. Clostridium botulinum is currently classified into four physiological groups (I to IV) (6, 16, 39) that differ widely in phenotypic properties and nucleic acid relatedness (16, 22). Although strains in the four groups are quite different physiologically and genetically, all have the common property of producing a characteristic botulinum neurotoxin (BoNT), of which seven serotypes of toxin (A through G) are recognized. When grown in media, foods, or the human intestine, C. botulinum produces toxin complexes where the BoNT is noncovalently linked with nontoxic proteins, including hemagglutinins (Hns) and a nontoxic nonhemagglutinating protein (NTNH) (10, 30, 32, 38). Strains in groups I and II, forming complexes of A, B, E, or F toxins, have been responsible for nearly all cases of human botulism (16). Although the nucleotide sequences of the genes encoding all seven serotypes of the BoNTs as well as tetanus toxin and some Hns and NTNHs have been determined (3, 4, 9, 13, 15, 17, 19, 21, 22a, 40, 41, 46), little is known regarding the regulation of expression of the various genes and the organization of the genes for the toxin complexes on host DNA. The genes for BoNT, Hn, and NTNH are closely linked on their host DNAs in C. botulinum types A (9), B (9), C (15, 18, 41), D (11), E (15), F (9), and G (50). In type C, the neurotoxin gene may form an operon with a gene for a nontoxic component (17), which operon may also exist in type A (48). The locations of the toxin-associated genes have been studied for certain serotypes. In types C and D, genes for BoNT, Hn, and NTNH are located on bacteriophages (tox ϩ bacteriophages) (11, 12, 41). The genes encoding the toxin complex are