Summary Objective Longitudinal study to test the validity and reliability of the Croatian version of the neck disability index (NDI-CRO) for use in patients. Methods Three groups were given NDI-CRO on two occasions, 48 h apart: acute whiplash neck injury group ( n  = 30), hospital physiotherapists—professional chronic neck pain group ( n  = 56) and control group ( n  = 65). To test validity, correlation between NDI-CRO and the pain VAS and PHQ‑9 questionnaire for depression was analyzed. Reliability testing was done using the test-retest experiment and item-total score correlation. Results Test-retest showed excellent correlation in all groups: whiplash 0.86, control 0.95 and physiotherapist 0.89 (Spearman r). Item-total score in the 3 analyzed groups showed positive correlation in all 10 categories, varying from 0.43–0.85. The NDI-CRO score showed significant difference between groups (median whiplash 38%, physiotherapist 12% and control 6%, p  < 0.05). Validity testing showed positive correlation between NDI and pain VAS (control r = 0.63, physiotherapist r = 0.68 and whiplash r = 0.83, p  < 0.05) and PHQ‑9 (control r = 0.49, physiotherapist r = 0.71 and whiplash r = 0.69, p  < 0.05). No correlation was found between NDI-CRO and age, gender or radiographic findings. In the whiplash injury group 75% of patients showed moderate to severe disability. The majority of physiotherapists showed mild to moderate disability due to neck pain while 40% showed no disability. In the control group two out of three people reported no disability. Conclusion The NDI-CRO is a valid index for measuring the degree of neck disability in people with acute and chronic neck pain and in control group. It is strongly correlated with pain VAS and PHQ‑9 index.