One hundred and one 5-year-old children were interviewed about a routine he
alth assessment carried out at school following delays of both 3 days and 1
year or 1 year only. Children were interviewed with prototypical medical i
tems and a doll (props), with verbal prompts only (verbal), or with drawing
(drawing). There was a decrease in both the amount and the accuracy of the
information children reported over the 1-year delay, but no effect of the
prior (3-day) interview. Children interviewed with props recalled more info
rmation than those asked to draw or interviewed with verbal prompts only, p
articularly at the long delay. Correct information was more likely to be re
peated across interviews than were errors, and, whereas information repeate
d across interviews was highly reliable, information introduced for the fir
st time after 1 year was not, particularly when children drew. These findin
gs have important implications in applied contexts such as when children ar
e called upon to provide testimony following very long delays. Copyright (C
) 2000 John Wiley & Sons, Ltd.