ABSTRACT Diagnostic device errors by health wearables cannot be avoided entirely, but they can have dramatic consequences for individuals, who are, consequently, deeply concerned and may refrain from using. However, it remains unclear how individuals assess and respond to potential diagnostic device errors when adopting health wearables. The present study unpacks this ‘black box’ using the elaboration likelihood model (ELM) to evaluate how potential diagnostic device errors translate into error perceptions, error processing, and behavioural reactions enacted through central and peripheral route cues. Based on a survey of 193 people with diabetes, we unveil that while peripheral cues are activated, most of the error processing is conducted consciously, strongly contributing to individuals’ attitude formation towards health wearables and their usage intention. These insights improve our theoretical understanding of user perceptions and responses to potentially erroneous health wearables. Furthermore, they guide suppliers in optimising their strategic product development and communication strategies.