Norflurazon was applied (2.24 kg/ha) in 1988 and 1989 to plots that co
ntained subsurface drains 1 m deep. After soybeans were planted, runof
f losses of the herbicide were measured each season for about 300 days
. The 1988-1989 season was atypical in that rainfall was insufficient
to produce runoff until 171 days after application. From day 171 until
day 302, 61 mm of runoff removed 1.5 g/ha (0.07 % of application) of
herbicide from the plots. In 1989-1990, runoff was measured for 301 da
ys after application. In this period 50.7 g/ha norflurazon (2.3 % of t
he amount applied) was lost from the plots in runoff. The highest conc
entration in runoff in the first study was 6.6 mug/L on day 171. The h
ighest concentration, 16.5 mug/L, for the second season was also measu
red during the season's first runoff event, on day 22. The studies pro
vided good examples of the direct relationship between time after appl
ication of runoff-producing rainfall and the magnitude of pesticide lo
ss in runoff.