Background. Two recent accounts of the use of causal criteria make opp
osite claims: that criteria should be used more often to avoid bias in
assessments of weak associations and, in direct contrast, that criter
ia are scientifically invalid. Methods. A recent review of the current
practice of causal inference in epidemiology, as well as some more th
eoretical concerns, reveals errors in the two claims. Results. In prac
tice, epidemiologists often use the criteria of consistency, strength,
dose-response, and biological plausibility, but not often temporality
, when judging weak associations. These criteria are used for causal a
ssessments as well as for making public health recommendations. In the
ory, causal criteria can be used to either refute or predict causal ef
fects. Conclusion. Research on causal inference methodology should be
encouraged, including research on underlying theory, methodology, and
additional systematic descriptions of how causal inference is practise
d. Specific research questions include: to what extent can consensus b
e achieved on definitions and accompanying rules of inference for crit
eria, the relationship of meta-analysis to the criterion of consistenc
y, and the interrelationships of criteria such as consistency, strengt
h of association, and biological plausibility.