Background <p>Patients with disorders of consciousness (DoC) are unable to partake in the decision making process concerning their treatment. In the process of medical decision-making, which in DoC often concerns life-and-death decisions, surrogates and healthcare professionals may try to reconstruct the treatment preferences of these patients. We aimed to identify which values and criteria have been used in various national care contexts to reconstruct the treatment preferences of incapacitated DoC patients and how reconstruction is conducted in practice.</p> Methods <p>This is a systematic review of the of conceptual and empirical ethical literature. A search was performed in seven databases (Pubmed, Web of Science, Embase, the Cochrane Library, CINAHL, PsychINFO, and Sociological Abstracts). We used thematic analysis to identify values and criteria for reconstruction of treatment preferences.</p> Results <p>8.591 records were screened. In total, 17 conceptual studies and 13 empirical studies on preference reconstruction were included. We identified three normative-epistemic viewpoints on reconstruction of treatment preferences: a correspondence view which aims to respect personal autonomy and prioritizes the accuracy of reconstructed preferences; a coherence view which aims to respect personal identity and prioritizes the consistency of the preferences with the patient’s lifeworld; and a communitarian view which aims to respect community and prioritizes the ongoing relation of a patient with family and friends. These views diverge on the problem of what makes for a good process of preference reconstruction. Additionally, treatment preferences of patients in DoC are inferred based on either past oral statements or on observations of patients’ current behavior. The criteria that guide reconstructive efforts may evolve, especially when patients improve from UWS to MCS and when reconstructed preferences based on past statements and patient’s current inferred psychological mental states steer treatment in mutually exclusive directions. There is no current standard approach to reconstructing treatment preferences in incapacitated DoC patients.</p> Conclusions <p>We recommend physicians to ask diversely formulated questions that stimulate surrogates towards giving multiple and rich answers. Simultaneously, physicians are advised not to overly test a surrogate’s testimony because this may lead to an erosion of trust.</p>