We examine whether propensity to participate in a web-CAPI mixed-mode survey is influenced by being contacted by email in addition to mail. In panel surveys, researchers can ask at each wave for an email address, but there is little evidence regarding the value of doing so. Using data from a large sample with an experimental design (IP5) we find that using a respondent-supplied email address to send additional invites and reminders does not affect response rates compared to using mailed invites and reminders alone, but results in more responses by web rather than CAPI and hence lowers survey costs. We find no evidence that these results depend on time spent in the sample.