Background and aims Neck pain can impair perception of cervical movement, but how this is affected by attention is unknown. In this study, the effects of experimental neck pain on head repositioning accuracy during standardized head movements were investigated. Methods Experimental neck pain was induced by injecting hypertonic saline into the right splenius capitis muscle in 28 healthy participants (12 women). Isotonic saline was used as control. Participants were blindfolded while performing standardized head movements from neutral (start) to either right-rotation, left-rotation, flexion or extension, then back to neutral (end). Movements were triplicated for each direction, separated by 5-s, and performed with or without a cognitive task at baseline, immediately after the injection, and 5-min after pain disappeared. Repositioning accuracy was assessed by 3-dimensional recordings of head movement and defined as the difference between start and end position. Participants were grouped into most/least accurate based on a median split of head repositioning accuracy for each movement direction at baseline without the cognitive task. Results The most accurate group got less accurate following hypertonic injection during right-rotation without a cognitive task, compared with the least accurate group and the isotonic condition (p < 0.01). No group difference was found when testing head repositioning accuracy while the participants where distracted by the cognitive task. Conclusions Experimental neck pain alters head repositioning accuracy in healthy participants, but only in those who are most accurate at baseline. Interestingly, this impairment was no longer present when a cognitive task was added to the head repositioning accuracy test. Implications The results adds to our understanding of what factor may influence the head repositioning accuracy test when used in clinical practice and thereby how the results should be interpreted.