A combined telephone contact-mail booklet-telephone interview of California and New England households regarding their willingness to pay for fire management in California and Oregon's old-growth forests was performed to test hypotheses regarding the spatial extent of the public goods market. Using a multiple-bounded contingent valuation question, the study found that New England households' annual willingness to pay for the California and Oregon programs was statistically different from zero. This analysis points out that households receive benefits from fire protection of old-growth forests in states other than their own. In this case study, limiting the survey sample to state residents where the National Forest is located would reflect about 20% of the national benefits. However, using resident values as a proxy for nonresidents would overstate the national benefits by 75%, since the values per household are significantly different. This finding suggests more emphasis in future surveys on selecting an institutionally and economically relevant sample frame rather than an expedient one.