Maintaining participant engagement in longitudinal surveys has been a key focus of survey research, and has implications for the quality of response and cost of administration. This paper presents new research measuring the impact of the design of between-wave keeping-in-touch mailings on response to the mailing and subsequent wave of a longitudinal survey. Three design attributes of the mailings were randomly implemented: the form of response request (whether respondents were asked to respond only if their address had changed, or in all cases to confirm or update their address); the newsletter included with the mailing (contrasting a newsletter with content tailored to respondent characteristics with a general newsletter and no newsletter); and the outgoing postage used (stamped or franked). The experiments were fielded on a new longitudinal study, the English and Welsh Civil and Social Justice Panel Survey (CSJPS), and took place between waves one and two. Fieldwork for both waves was conducted by Ipsos MORI face-to-face interviewers. Our main finding was that the tailored newsletter was associated with a significant increase in the wave-two response rate. However, in relation to response to the request, the tailored newsletter, or sending no newsletter at all, were equally effective at inducing response, and significantly better than the general newsletter. We also found that, in relation to the form of request, the ‘change of address’ request was as effective as the more costly ‘confirmation’ request. Findings are discussed with reference to the design of keeping-in-touch mailings for longitudinal surveys.