Several herbicides including thiobencarb (benthiocarb, S-4-chlorobenzyl diethylthiocarbamate) and chlornitrofen (CNP, 4-nitrophenyl 2, 4, 6-trichlorophenyl ether) were studied on their behavior and transformation in soils after application to paddy fields for the control of weeds.(1) The soils for analysis were sampled at definite time intervals, to definite surface area and to definite depth by a cylinder as shown in Fig. 1. The fresh weight, dry matter weight and bulk density were recorded to calculate the ratio of residue to the amount of herbicides applied.(2) Large amounts of elemental sulfur which were formed in the soils under the reduced condition, were extracted with organic solvents along with the herbicide in the course of thiobencarb analysis. The disturbance due to elemental sulfur in the gas chromatograms was easily eliminated by shaking the hexane extract with the aqueous solution of tetrabutylammonium sulfite.(3) Thiobencarb in soils was slowly degraded under the paddy field conditions. The amount of residues in autumn was lower in Miyazaki, and higher in Hokkaido than in Saitama, reflecting the effect of temperature on the degradation process.(4) By applying an analytical method for thiobencarb, a new metabolite, dechlorinated thiobencarb (Fig. 7) was detected in the paddy soil. This compound was highly toxic to rice plants both in the germinating stage and the 3-leaf stage, and found to be the substance responsible for the dwarfing of rice plants. It appeared in soil at 25 days after the application of thiobencarb, reaching a peak at 35-45 days, and then rapidly disappeared (Fig. 8). This occurrence coresponded to the appearance of the dwarfing symptoms on rice plants. The reductive dechlorination of thiobencarb took place in the upper layer of the paddy soil, when the soil was applied a large quantity of organic matter including rice straws, kept for a long period of time under the flooded conditions, and also related to the activity of certain microorganisms. Recently, it has been demonstrated that this reaction can be selectively inhibited by the combination with safeners such as BNA-80.(5) CNP in the paddy soil was rapidly reduced to the amino derivative which persisted many years in the form of residues bound to humic substances. The total CNP-amino content was determined by extraction with organic solvents, after heating the soil in 5N KOH solution with sodium sulfide for 20 hours in a boiling water bath. From the content of the total CNP-amino, the CNP-amino content was calculated by the subtraction of that formedd from CNP (nitro) in the process of the alkaline treatment.(6) From 1973 through 1975, soil samples were collected in many locations in Japan after the harvest of rice grains from paddy fields treated with CNP. Generally speaking, the contents of CNP-amino in the soils were very high. As shown in Fig. 12, the ratios of residues of CNP (nitro+amino) to CNP applied annually were scattered within wide ranges. Values exceeding 100% in many soil samples in Hokkaido, suggested the presence of cumulative residues. Besides the temperature, factors such as clay contents and duration of flooding periods influenced the amount of residues, suggesting differences in microbial degradation.(7) In the paddy plot in Tsukuba to which a large quantity of rice straws and grains had been incorporated, the chlorine atom at the para position of CNP (2, 4, 6-trichlorophenyl compound) was replaced with a hydrogen atom, and thereafter the remaining chlorine atoms were replaced in turn (Fig. 14). In the conventionally cultivated paddy fields, sometimes 2, 4-dichlorophenyl compound which was replaced at the ortho position of CNP, was detected beside the 2, 6-dichlorophenyl compound.(8) Thiobencarb, CNP and the derivatives mentioned previously were to some extent taken up by rice plants, but they hardly translocated into the plant parts, and were never detected in rice grains.