Question: Does change in impairments within and between the first two manual therapy treatments predict change in activity limitations by the end of treatment in patients with subacute neck symptoms? Design: Longitudinal, observational study. Participants: 29 people with neck pain for more than two weeks who subsequently received = three treatments. Outcome measures: Impairments measured were active neck ROM in six directions (total ROM), most limited direction of ROM (limited ROM), pain intensity, and pain location. Activity limitations were measured using the Neck Disability Index and the Patient Specific Functional Scale. Patients' perceptions of change were measured using the Global Perceived Effect Scale. Impairments and patients' perceptions were measured before and after the first two treatments and before the final treatment whereas activity limitations were measured only before the first and last treatments. Results: All measures improved by the end of treatment. Between-treatment change in limited ROM predicted change in limited ROM (rs 2 = 0.53 and 0.57) and total ROM (rs2 = 0.26) by the end of treatment. Within- and between-treatment change in pain location predicted change in pain location (rs 2 = 0.24, 0.27,0.28, and 0.57) by the end of treatment. No significant relationships were found between change in any impairments in the first two treatments and change in activity limitations by the end of treatment. Conclusions: Change in impairments predicts change in the same impairment by the end of treatment, but not in other impairments or activity limitations. It is recommended that the reassessments used to guide and refine treatment be individualised and related to the specific goals for that patient.