Insect populations succumb to a variety of infections caused by pathogenic microorganisms (bacteria, fungi, Protozoa) and viruses. The narrow host-range of many of these agents makes them natural candidates for use within integrated pest management systems. Some, such as Bacillus thuringiensis and several baculoviruses, may be applied to crops at regular intervals as microbial pesticides, achieving short-term control of a pest population. Longer-term suppression of insect populations requires some degree of persistence of the pathogen in the target host population. Examples of sustained, natural, insect population regulation by microorganisms are rare; regulation demands stable ecosystems and a capacity for the pathogen to spread. We cannot ignore the fact that many of the microbial pathogens available today fail to meet the expectations of an agricultural industry used to the rapid and broad-spectrum pest knockdown achieved by many chemical pesticides. Despite the many advantages to be gained in selective pest management from the use of naturally occurring strains of insect pathogens, much recent attention has focused on the improvement of strains by genetic manipulation. Significant advances have already been made in the manipulation of bacterial and viral pathogens to increase virulence and modify host range. The environmental persistence of the insect-pathogenic toxin of B. thuringiensis has also been extended by inserting the toxin gene into other bacterial hosts and plants. Exciting future opportunities for biological control may be created by such strategies. However, to make responsible use of these manipulated organisms we must understand more about their long-term impact on insect populations and the environment. Such information should come not only from detailed ecological studies of the host-pathogen interaction but also from laboratory and field studies of the frequency and consequences of genetic exchange between modified strains and naturally occurring microorganisms.