Background Health anxiety (HA) by proxy is described as parents’ obsessive worries that their child is severely ill although this is not medically confirmed. Research on HA by proxy suggests that it is highly distressing for the parent and that the child may be at risk of developing maladaptive symptom coping strategies. No targeted treatment for this group exists. We developed PROXY, an 8-week psychological internet-delivered treatment for parents with HA by proxy. The treatment components of PROXY are informed by cognitive behavioral therapy as well as acceptance and commitment therapy, and it was developed in collaboration with parents experiencing HA by proxy and clinical experts. Objective This paper describes the protocol for a study investigating the potential effects of PROXY on parents’ worries about their children’s health using a single-case experimental design (SCED). Methods Five parents clinically evaluated as experiencing HA by proxy will be included. A replicated randomized SCED study will be conducted in which each participant will be randomized to receive treatment after a baseline period of between 7 and 26 days (phase A). The treatment phase duration is 8 weeks for all participants (phase B), followed by a follow-up phase lasting between 14 and 33 days (phase C), ensuring that all participants remain in the study for 96 days. Participants will report daily anxiety level by SMS text message throughout the study. They will also answer self-report questionnaires, including questions on HA by proxy and own HA, 4 times during the study. Data will be submitted to structured visual analysis to inspect anxiety level, the variability of anxiety, trends, the overlap of data points among phases, when effects occur, and the consistency of data patterns across participants. Furthermore, randomization tests will be conducted for each participant to test the null hypothesis that PROXY will have no effect on participants’ anxiety. Results The recruitment of parents began in June 2022. As of March 2, 2023, a total of 4 parents have been included in the study. Data collection was expected to cease in April 2023. Conclusions To the best of our knowledge, this protocol describes the only study of treatment for HA by proxy. As the prevalence of this condition is still unknown, a SCED was chosen because this method allows the inclusion of very few participants while still providing information on effects and treatment courses. Conducting the study with a replicated randomized phase design enables methodologically sound testing despite the inclusion of very few participants. The results will inform researchers on individual treatment courses and effects, which may direct future research in terms of the possible mechanisms of change, ideas for how to refine the treatment content, and how the treatment may be offered to patients in the future. Trial Registration ClinicalTrials.gov NCT04830605; https://clinicaltrials.gov/study/NCT04830605 International Registered Report Identifier (IRRID) DERR1-10.2196/46927