Relationships among; soil organic matter (SOM), the feeding activity of the soil-dwelling scarab larva, Costelytra zealandica (White) (Coleoptera: Melolonthidae), amber disease and soil numbers of its causal agent, Serratia entomophila (Enterobacteriaceae), were studied. C. zealandica larvae may cause severe damage to pasture roots and reduce ryegrass and clover production. Entomopathogenic strains of S. entomophila are being used in a biocontrol programme in New Zealand as they induce inhibition of larval feeding activity and larval death 1–3 months from the onset of infection. A mature (38-year-old) and a young (4-year-old) pasture from Canterbury, New Zealand, were chosen for this study. Both pastures had the same soil type, very similar agronomic practices but different ages and therefore SOM contents. Soil from both sites was collected and amounts of C and N of different fractions of SOM were measured. Organic matter was 7.59% and 4.41% in the mature and young pastures respectively. Total C and N in soil and most fractions of SOM were significantly higher in the mature pasture than in the young pasture. Total C and N present in cold and hot water extractions combined, representing the labile SOM fraction, showed 45% less C and 37% less N in the soil from the young pasture than in the soil from the mature pasture. In both soils, approximately 10% of the total C and N was present in the labile SOM fractions. Populations of soil-borne indigenous bacteria, S. entomophila, were estimated in fresh and air-dried samples from both pastures. Soils from the two pastures contained a significantly different number of S. entomophila cells. A more than ten times higher population was present in fresh soil from the mature pasture compared with fresh soil from the young pasture. However, levels of S. entomophila decreased significantly when the soil was air-dried for experimental purposes. An experiment was designed to explore the interactions occurring among SOM quantity and quality, S. entomophila performance as a biocontrol agent and simulated root feeding activity of early third instars of C. zealandica. Larvae were placed individually in pots containing air-dried soil from one or other of the two pastures ( n = 10 per treatment). The experimental pots were inoculated at three dose rates of a pathogenic strain of S. entomophila. The control pots were treated with nutrient broth free of S. entomophila. All pots were incubated in the laboratory under the same experimental conditions. Larval feeding activity was measured using pieces of carrot root. All pots were examined for infection, mortality, feeding rate and gain in live weight of each larva after 15 and 30 days. The mature pasture was associated with significantly higher amber disease and mortality than the young pasture. Also, a significantly higher gain in larval weight was observed in the young pasture soil, together with a higher rate of carrot consumption by larvae. It can be concluded that the content of labile SOM in both pastures may account for the differences observed in this study and could have a buffering effect on plant damage caused by soil-dwelling scarab larvae.