Many regulating ecosystem services exhibit non-excludability and non-rivalry characteristics akin to public goods. This makes it unlikely that such regulating ecosystem services will attract a price in the marketplace. Policymakers, therefore, find difficulties in determining a correct economic valuation for regulating ecosystem services, with potentially severe consequences for aligning policy for ecosystem service provision with public preferences for service delivery. This research used a choice modelling approach to estimate public preferences for delivery of two regulating ecosystem services; reductions in the flood risk to the British city of York (implemented by filling in drainage ditches in peat moorland further up the catchment), and a reduction in CO2 emissions (implemented by planting Poplar trees for biomass either locally in the study area or elsewhere in the UK). The choice experiment presented respondents with options providing different levels of CO2 reduction, implemented locally or nationally, and different levels of flood risk reduction, at a range of different prices, with a tax as the payment vehicle. Subsequent analysis with latent class (LC) models showed considerable heterogeneity of preference among respondents, particularly with regard to the location of tree planting for CO2 reduction. One LC segment showed a significant preference for achieving a high level CO2 reduction through national, as opposed to local tree planting, potentially indicating the existence of NIMBYism. The same LC segment displayed a preference for moderate reductions in the flood risk to York, even though this would deliver no personal benefit for a substantial proportion of the segment. This could, therefore, indicate altruistic behaviour towards geographically proximate beneficiaries.