On the basis of the known cry gene sequences of Bacillus thuringiensis, three sets of primers were designed from four conserved blocks found in the delta-endotoxin-coding region. The primer pairs designed amplify the regions between blocks 1 and 5, 2 and 5, and 1 and 4. In silico analyses indicated that 100% of the known three-domain cry gene sequences can be amplified by these sets of primers. To test their ability to amplify known and unknown cry gene sequences, 27 strains from the CINVESTAV (LBIT series) collection showing atypical crystal morphology were selected. Their DNA was used as the template with the new primer system, and after a systematic amplification and sequencing of the amplicons, each strain showed one or more cry-related sequences, totaling 54 different sequences harbored by the 27 strains. Seven sequences were selected on the basis of their low level of identity to the known cry sequences, and once cloning and sequencing of the complete open reading frames were done, three new cry-type genes (primary ranks) were identified and the toxins that they encode were designated Cry57Aa1, Cry58Aa1, and Cry59Aa1 by the B. thuringiensis Toxin Nomenclature Committee. The rest of the seven sequences were classified Cry8Ka2, Cry8-like, Cry20Ba1, and Cry1Ma1 by the committee. The crystal morphology of the selected strains and analysis of the new Cry protein sequences showed interesting peculiarities.Bacillus thuringiensis has been safely used for the control of insect pests within the orders Lepidoptera, Coleoptera, and Diptera for the last 50 years (19). It is an aerobic, Grampositive bacterium whose insecticidal activity is based on the presence of parasporal crystalline inclusions formed during the sporulation process. These parasporal bodies or crystals are assembled by the so-called Cry proteins, expressed by the cry genes (21).B. thuringiensis shows great variability, as has been demonstrated by the huge number of strains isolated around the world (19), by the number of serotypes known to date (a total of 84) (20), and by the great number of different cry gene sequences accumulated so far (a total of 492) (8), as well as by the number of molecular characterization tools that have been developed, such as sequencing of the flagellin gene and of the gyrB and aroE genes, the band patterns from repetitive extragenic palindromic-PCR analyses, and the plasmid patterns, among others (17,18,25,27), all indicating the great variability within this species.In spite of this variability, some similarity has been found in the three-domain Cry proteins, evidenced by the presence of three conserved domains in the tertiary structure of the active toxin (delta-endotoxin), even from Cry toxins showing low levels of identity (i.e., the Cry1-type, Cry3-type, and Cry4-type toxins): one domain with a bundle of ␣ helices and two domains of  sheets (3). This uniformity is, at least in part, a reflection of the five conserved blocks in the gene structure, which are present in almost all the cry genes (10).The great number of...