Introduction Patients suffering from remaining disability after anterior cervical decompression and fusion (ACDF) surgery for cervical disc disease may be prescribed physical activity (PPA) or neck-specific exercises (NSEs). Currently, we lack data for the success of either approach. There is also a knowledge gap concerning the use of internet-based care for cervical disc disease. The scarcity of these data, and the high proportion of patients with various degrees of incapacity following ACDF, warrant increased efforts to investigate and improve cost-effective rehabilitation. The objective is to compare the effectiveness of a structured, internet-based NSE programme, versus PPA following ACDF surgery. Methods and analysis This is a prospective, randomised, multicentre study that includes 140 patients with remaining disability (>= 30% on the Neck Disability Index, NDI) following ACDF for radiculopathy due to cervical disc disease. Patient recruitment occurs following attendance at routine clinical appointments, scheduled at 3 months postsurgery. Patients are then randomised to one of two groups (70 patients/group) for a 3-month treatment programme/period of either internet-based NSE or PPA. Questionnaires on background data, pain and discomfort, physical and mental capacity, satisfaction with care, and health and workplace factors are completed, along with physical measures of neck-related function conducted by independent test leaders blinded to randomisation. Measures are collected at inclusion, after the 3-month treatments (end of treatment) and at a 2-year follow-up. Radiography will be completed at the 2-year follow-up. Preoperative data will be collected from the Swedish Spine Registry. Data on healthcare consumption, drug use and sick leave will be requested from the relevant national registers. Ethical considerations This study was approved by the Regional Ethical Review Board in Linkoping Ref. 2016/283-31 and 2017/91-32. The scientists are independent with no commercial ties. Patients are recruited after providing written informed consent. Patient data are presented at group level such that no connection to any individual can be made. All data are anonymised when reported, and subject to the Swedish Official Secrets Health Acts. The test leaders are independent and blinded for randomisation. Exercises, both general and neck-specific, have been used extensively in clinical practice and we anticipate no harm from their implementation other than a risk of muscle soreness. Both randomisation groups will receive care that is expected to relieve pain, although the group receiving NSE is expected to demonstrate a greater and more cost-effective improvement versu s the PPA group. Any significant harm or unintended effects in each group will be collected by the test leaders. All questionnaires and test materials are coded by the research group, with code lists stored in locked, fireproof file cabinets, housed at the university in a room with controlled (card-based) access. Only individuals in receipt of a unique website address posted by the researchers can access the programme; patients can neither communicate with each other nor with caregivers via the programme. Study participation might lead to improved rehabilitation versus non-participation, and might therefore be of benefit. The results of this study should also contribute to more effective and flexible rehabilitation, shorter waiting times, lower costs and the possibility to implement our findings on a wider level. Dissemination If effective, the protocols used in this study can be implemented in existing healthcare structures. The results of the study will be presented in scientific journals and popular science magazines of relevance to health. The findings will also be presented at local, regional, national and international conferences and meetings, as well as in the education of university students and at public lectures. Information about the results will be communicated to the general population in cooperation with patient organisations and the media.