Liquid poultry manure was applied to three forest sites at 35, 70, and 225 metric tonnes (1 tonne = 1000 kg) per hectare (1 ha = 10 000 m2) for 1 to 3 years. Two of the sites were white pine plantations, one dry, one moist, and the third site was mixed hardwoods. Over a period of 35 months the total bacterial population in litter and in soil, in both control and manure plots, was studied as well as the ability of the population to produce extracellular proteases, cellulases, lipases, pectinases, and amylases. Total bacterial counts in litter tended to be higher in the autumn and lower in the spring. Litter always contained about 10-fold more bacteria than soil and the number of bacteria in soil fluctuated little with season or treatment. In white pine litter, bacteria able to excrete protease were more numerous than other enzyme-producing bacteria. More bacteria able to excrete lipases were found on moist than on dry sites, especially on manured plots. At the end of the 35-month test period, all bacterial counts in manured plots had returned nearly to the original level of the control. Statistical analysis revealed significant differences in total bacterial count and counts of specific enzyme producers between manured and control plots. Little difference was seen among rates of manure application. Adding large amounts of poultry manure to these forest plots did not greatly change the bacterial ecology of the soil or litter.