Information on respondents' perceptions about survey consequentiality is typically collected close to the end of the survey, following the preference elicitation. We inquire whether—and if so, how—the location and a repetition of a consequentiality perception elicitation question matter for stated consequentiality perceptions and for stated preferences. To that end, we use data from a discrete choice experiment survey conducted in Germany, in which respondents evaluated a project of expanding urban green areas. The survey involved two treatments: in one, respondents were asked about their consequentiality perceptions only after completing all preference elicitation tasks; in the other, respondents' consequentiality perceptions were elicited twice, that is, before and after the preference elicitation tasks. Based on ordered logit models, we find that stated consequentiality is stronger (i) when respondents are asked about consequentiality perceptions twice (that is, before and after preference elicitation tasks) rather than when asked only once (that is, after the tasks); and (ii) when the perceptions are inquired before the preference elicitation tasks rather than when inquired after the tasks. The latter finding emerges in both within-sample and across-sample comparisons. Mixed logit willingness-to-pay-space models reveal that stated willingness-to-pay values for project attributes significantly differ depending on whether the consequentiality elicitation question is asked before the preference elicitation or not. Furthermore, the consequentiality perceptions appear to significantly affect stated willingness-to-pay values, and this influence differs depending on whether the consequentiality elicitation question is asked before or after preference elicitation tasks. The findings of differences in the impact of the way consequentiality perceptions are elicited on willingness-to-pay values of attributes are observed for each of four German cities in which the survey was conducted. This evidence suggests that when willingness-to-pay values are corrected based on stated consequentiality perceptions—as often done in recent stated preference studies—the values may be sensitive to the way the consequentiality perceptions are elicited.