The persistence and disappearance (washoff or dryfall) of methoxychlor [2,2-bis(p-methoxyphenyl)-1,1,1-trichloroethane] from mature soybean [Glycine max (L.) Merrill] foliage was investigated in a small field plot study under natural rainfall conditions in 1977 and 1978. Residue analyses were conducted using whole plant samples. Methoxychlor washoff rate was 8+/- 4% of the amount on plants (prior to rain) per centimeter of rainfall, regardless of time after application. Total seasonal washoff for 1978 accounted for 33.5% of the applied pesticide; however, 30.5% of the total loss was removed by washoff on the second day after application. Dryfall or dislodgeable residue accounted for less than 1% of the amount applied. The amount of dryfall was significantly greater in plots entered by workers than in those where entry was avoided. More than 19% of the applied methoxychlor was lost as a result of through-fall to the ground during application to the plots. Statistical analyses indicated that within-sample variation for mechanical chopping of plant samples was significantly smaller at the 5% level than for a hand chopping method. Results from this study will be useful in defining research objectives for the development of algorithms to describe the behavior of foliar-applied compounds. Such algorithms are necessary for estimating runoff losses of insecticides to water bodies.