To understand pesticide runoff due to surface water flow from sloped cropland (6-6.5) with planted cabbages, runoff model tests with artificial rainfall were performed in different sized blocks of land. Immunochemical determinations using commercially available immunoassay kits were applied for pesticides in surface water. One hour after application of each 400 ppm emulsion of chlorothalonil (TPN) and diazinon, artificial rainfalls of 60 mm/hr for 10 and 23 min were induced to two small fields (blocks A and B: 6. 3 m2 each) and the rainfall of 18. 7 mm/hr for 30 min and the next day 25. 2 mm/hr for 10 min were also induced to a large field (block C: 460m2). The ratios of runoff water recovered were 1. 2% each in the blocks A and B, and 6. 0% from the first rainfall and 5. 8% from the second rainfall in the block C. The relative pesticide runoff was 0. 14-0. 62% for TPN and 0. 06-0. 39% for diazinon. When the results of analyses by both immunoassay and conventional GC were compared, they seemed to be roughly identical in similarly collected samples. Furthermore, deposits of pesticides on cabbages and soil surfaces were analyzed by immunoassay and the mass balance of each pesticide was estimated before and after rainfall in the test fields. TPN seemed to deposit better on cabbages and soil surfaces than did diazinon. Though in the block C almost half of the both pesticides initially deposited on the cabbages was washed off during the first rainfall, additional wash-off from the cabbages was observed only a little during the second rainfall. Before and after the second rainfall almost the same concentrations of each pesticide were found in the soil surface layer.