Associations between fluctuation of treatment need for temporomandibular disorders (TMD) and age, gender, stress, and diagnostic subgrouping were analyzed in a 2-year follow-up of 391 subjects. All the studied factors were significantly associated with the treatment need for TMD at all examinations. The diagnostic subgroup (TMD arthro, TMD myo, TMD comb, or non-classified) at base line was significantly associated with the fluctuation of the treatment need for TMD also during the follow-up, but age, gender, and stress score were not. In the subgroup needing active treatment for TMD at least once during the follow-up (n = 65), the stress score did not show statistically significant covariation with the treatment need. The diagnostic subgrouping of these 65 subjects at the second and third examination at 12-month intervals did not show any association with the subgrouping at base line or with any studied variable. Detailed descriptive diagnostics may serve well in treatment planning but do not necessarily help us in understanding the nature of TMD.