Conventional and molecular methods for the improved detection of B. anthracis in environmental material were studied. A system was developed which combines a selective enrichment broth with two-phase concentration using polyethylene glycol and potassium phosphate to form the two immiscible phases. The enrichment broth alone, based on polymyxin B sulphate, lysozyme, EDTA and thallous acetate, which comprise PLET agar (PLETA), allowed the selective recovery of B. anthracis from a mixture of B. anihracis and closely related B. cereus. When soil was added to the broth, however, B. anthracis was rapidly overgrown by other naturally occurring Bacillus species. Recovery of B. anlhracis was improved by using semi-solid PLET broth or by adding chelating agents or the monobactam antibiotic aztreonam to the broth. The combination of chemicals required for optimal recovery of B. anthracis varied according to the composition of the soil. Use of the two-phase concentration system showed that in soil B. anthracis spores are generally attached to soil particles and need to be separated before they can be concentrated. Separation was achieved by pre-soaking the soil in water. The sensitivity of standard PLETA is approximately 5- 50 spores per gram of soil depending on the sample composition. The system finally recommended for the most reliable and sensitive detection of spores in soil achieved an average 25 fold greater sensitivity than PLETA. Further enrichment of the B. anthrcicis concentrate obtained using the optimised enrichment method allowed the PCR detection of B. cinthracis DNA. The sensitivity of the PCR was affected by the composition of the soil. In the absence of inhibition the PCR detection limit was approximately 10 - 100 spores per gram. A multiplex PCR was developed which targets DNA from pXO1, pX02 and the B. anthracis chromosome. The PCR allowed the rapid identification of colonies suspected of being B. anthracis. In addition to being essential for the definitive identification of B. anthracis, the ability to determine the presence of virulence plasmids in B. anthracis has reduced the need to use animals for virulence tests. Attenuated pX01⁺/pX02^- or pX01^-/pX02^- strains of B. anthracis are occasionally found in the environment. Naturally occurring pX01^-/pX02^- derivatives have not been isolated. No other plasmid DNA has been identified in B. anthracis. To examine the nature of the stability of pX01 and pX02 in B. anthracis, the effect of selective pressure for non-indigenous plasmid DNA that had been introduced into B. nnthracis was studied. A plasmid based on the minimal replicon of pAMß 1 (pAEX-5E, 5.8. kb) was found to be stable in pX01⁺/pX02^- and pX01^-/pX02⁺ derivatives of B. anthracis for more than 100 generations of growth. In the pX01⁺/pX02^- and pX01⁺/pX02^- derivatives of B. anthracis, pAEX-5E was expelled within 105 culture generations. Loss of pAEX-5E was most rapid in the pX01⁺/pX02^- derivative. Plasmids pXOI and pX02 both remained stable under selection pressure for pAEX-5E, and in the pX01⁺/pX02^- derivative retention of pX01 led to a reduction in growth rate. This indicates that, in the absence of pX02, B. anlhracis will endure a significant metabolic compromise in order to retain pX01. This study has provided extensive new information about the selective recovery of B. anthracis. in environmental material and novel'data about the stability of identity plasmids pXO1 and pX02 under selective pressure for non-indigenous plasmid DNA.