ABSTRACT:Two sites from a humid tropical environment were studied with respect to soil water repellency caused by hydrocarbon contamination. Samples were analyzed for water repellency (molarity ethanol droplet method), total petroleum hydrocarbons, acute toxicity (Microtox) and field capacity. At both sites, water absorption times were logarithmically related to the molarity ethanol drop value (R > 0.95). In a sandy soil collected from an old separation battery which had been bioremediated, field capacity was strongly related to hydrocarbon concentration (R = 0.998); and at 10,000 mg/kg the calculated field capacity was only 75 % of the baseline. Water repellency was related to hydrocarbon concentration asymptotically and plant growth limiting values (severity > 3.0) were observed at low concentrations (2,400 mg/kg), even though toxicity was at, or below background levels. Bioremediated soil at this site had hydrocarbon concentrations only 1,300 ppm above background, but had extreme water repellency (severity = 4.6 -4.7). Soil water repellency was also measured in a clayey, organic rich floodable soil, in a multiple pipeline right-of-way colonized by water tolerant pasture and cattails. Water repellency was associated with total petroleum hydrocarbon concentration (R = 0.962), but was not related to field capacity or toxicity. In this low-lying site, the water repellency observed in the laboratory is probably not representative of field conditions: samples taken at the end of the ten week dry season (and only four days before the first rains) showed ample moisture (> 80 % field capacity).