Chemoprevention can be defined as the use of chemical compounds or med
icines to prevent the occurrence of precancerous lesions (markers) or
to slow down or revert the progression of clinically established disea
se. The use of randomized trial design is considered the gold standard
for evaluating the preventive value of chemicals against cancer, sinc
e they control for confounding and avoid information bias. The princip
al school in relation to cancer control through chemoprevention is bas
ed on studies of cancer and diet. Initially, ecological studies set th
e cornerstone, but later case-control studies supported the hypothesis
of an inverse association between foods and cancer risk (principally
epithelial), suggesting that determined micronutrients participate as
protection in this process. Other studies include specific chemical an
alyses, which have potential problems that could lead to erroneous con
clusions, such as sample and measurement errors. During this decade ra
ndomized intervention trials have been carried out to test this hypoth
esis, but conclusions have been so diverse and the designs used have b
een so different in terms of levels of exposure, that consistent concl
usions are not possible. We can conclude that using studies with rando
mized, double-blind, controlled designs is interesting, but problems r
emain to be solved, including: agent selection, the design to be chose
n, and especially the balance between benefits sought and secondary ef
fects, including cost-effectiveness, since some chemicals cannot compe
te with other preventive or therapeutic measures.