Petroleum hydrocarbon contaminants in soil may leach to the groundwater and pose a threat to the quality of drinking water in Hawaii. This study was initiated to evaluate the competence of tropical bacteria, indigenous to Hawaii, to degrade petroleum hydrocarbons in soil when environmental factors are optimized to favor the growth of microorganisms. In a first experiment, baseline microbial populations were determined in five soils, then the soil was treated with either fertilizer or fertilizer and 3.000 mg/kg of diesel No. 2 and incubated for 20 days. Heterotrophic bacteria and phenanthrene-degrading bacteria were enumerated on plates, total bacteria were determined by microscopy (acridine-orange counts), and hexadecane-degraders were enumerated by the most-probable-number technique in multi-well plates. The double-layer-phenanthrene plates were modified and phenanthrene-degraders were found in all soils and were stimulated by the addition of diesel and fertilizer. Thus, Hawaii soils harbor microorganisms that can be active in the bioremediation of petroleum-contaminated soils. Of the 200 bacteria that were isolated from either the phenanthrene plates of the MPN-hexadecane wells, 70 could utilize as sole carbon source at least one of the following hydrocarbons: hexadecane, diesel, mineral oil, phenanthrene, and pyrene. Most isolates were specialized in their use of hydrocarbons. Of the 32 confirmed Gram-negative bacteria, 26 were identified by the Biolog??? system as belonging to the genera Pseudomonas, Sphingomonas, Acinetobacter, ,and Flavobacterium; the other 6 could not be identified. None of the 16 Gram-positive rod-shaped hydrocarbon degraders was indentified by the Biolog??? system. The second experiment was a time-course experiment (124 days) in a chronically contaminated soil (87 ppm of gasoline). This soil (#9) was spiked with 6,000 mg of diesel No. 2/kg, fertilized, and incubated in jars at 30??C and 40??C. The higher temperature was used to simulate the increase in temperature which may occur when bioremediation is conducted in a tent exposed to the sun under tropical conditions. Three rapid techniques (less expensive than gas chromatography) were evaluated to monitor the disappearance of hydrocarbons: gravimetry of a hexane extract to measure total petroleum hydrocarbons (TPH), an immunoassay to determine the concentrations of polycyclic aromatic hydrocarbons (PAH), and the Microtox??? assay to determine soil toxicity, The concentration of TPH and PAH decreased rapidly during the first month (25 and 10% left, respectively) and then declined very slowly in soil containing active microorganisms. Soil toxicity decreased with time. Similar results were obtained at 30??C and 40??C. The concentrations changed little in soil where the microorganisms had been killed with 0.5% HgCl2. CO2 evolution by the soil confirmed the growth of microorganisms in diesel No. 2. Determination of TPH by gravimetry and of PAH by immunoassay were more rapid, informative, and cost-efficient methods than the Microtox??? assay. Bacterial counts increased during the first 40 days, then remained stationary (total bacteria) or decreased (hydrocarbon degraders). In the third experiment, the possible enhancement of bioremediation of soil contaminated with diesel No. 2 and No. 6 (Bunker C) by inoculation with a versatile hydrocarbon-degrading bacterium (138) was examined. A clayey soil was contaminated with 6,000 mg/kg of either diesel fuel, limed, fertilized, seeded or not with bacterium 138, limed, fertilized, and incubated in jars at 30??C for 138 days. Poisoned controls (0.6% HgCl2) were used to determine the extent of hydrocarbon degradation due to microbial activity. A rapid biodegradation of TPH (75% in 14 days) occurred in soil contaminated with diesel No. 2, regardless of bacterial seeding. Biodegradation of PAH was more gradual but reached 90% by day 98 in both seeded and unseeded treatments. Inoculation (5 x 10^7 bacteria/g of soil) increased the counts of phenanthrene-degrading bacteria and of microorganisms capable of utilizing hexadecane and diesel No. 2. The counts of total bacteria and CO2 evolution were not increased by seeding. In soil contaminated with diesel No. 6, the measurements of TPH and PAH were more variable due to the uneven distribution of the product. The extent of the bioremediation of diesel No. 6 in this soil is unclear from these measurements. The counts of total bacteria remained unchanged after the addition of diesel No. 6. However the counts of the indigenous phenanthrene-degrading bacteria increased dramatically (4 log units) during the first 54 days whereas the level of the seeded bacteria remained stable. The counts of mineral oil degraders decreased steadily possibly due to the toxicity of diesel No. 6. A small effect of seeding was visible (54 to 138 days) in the amount of CO2 evolved by the soil contaminated with diesel No. 6. In conclusion, diesel No. 2 was readily degraded by soil microorganisms while diesel No. 6 was more refractory. In this soil and with the Gram-positive bacterium we used there was no detectable effect of inoculation on the extent of bioremediation. Other soils and inoculants should be examined before definitive conclusions can be drawn.