In contingent valuation, embedding refers to the solicitation of willi
ngness to pay for a good that is valued as a component of a larger goo
d. This study addressed one of the criticisms of previous embedding st
udies, that the good was inadequately described to respondents. We tes
ted for the effect of embedding on willingness to pay for natural area
protection under three information levels, and found that amount of i
nformation had little impact on the effect of embedding on willingness
to pay-under all three information conditions, embedding significantl
y lowered willingness to pay. If this and other embedding studies can
be interpreted as indicating that many people consider related public
goods to be close substitutes, then embedding studies demonstrate the
considerable sensitivity of respondents to information about substitut
es. Successful use of contingent valuation to value public goods relie
s on agreement about how to present information about substitutes.