Background: While cross-sectional associations of inflammatory markers interleukin-6 (IL-6) and C-reactive protein with major depressive disorder are well established, evidence for longitudinal associations mostly comes from studies on depression symptoms, not diagnoses. This study examined cross-sectional and bidirectional longitudinal associations between depression diagnosis and symptoms in an adult sample over a 6-year period. Methods: Data were obtained from the baseline (n = 2416) and 2- and 6-year follow-up assessments (n = 1925 and n = 1924, respectively) of the Netherlands Study of Depression and Anxiety. C-reactive protein and IL-6 were assessed at each wave, as were the Composite International Diagnostic Interview and Inventory of Depressive Symptomatology. Linear mixed models and generalized estimating equation models with a binomial distribution were used to study longitudinal associations between depression and inflammation and vice versa. Results: There was a consistent cross-sectional association between current depressive disorder (vs. no current disorder) and symptoms with IL-6 across all follow-up measurements (Cohen's d depression diagnosis = 0.06, p =.017; B standardized Inventory of Depressive Symptomatology = 0.029, SE = 0.011, p =.008). In longitudinal analyses, higher IL-6 levels predicted subsequent chronic course in those with a diagnosis at baseline in women but not in men (odds ratio women = 1.13, 95% confidence interval = 1.04–1.23), and both depressive disorder and high severity predicted higher IL-6 levels at the subsequent follow-up (p values <.01). In contrast, C-reactive protein was not associated with current depression in cross-sectional and longitudinal analyses. Conclusions: In this longitudinal study, cross-sectional and bidirectional longitudinal associations were found between depression and IL-6 levels. This underlines the importance of targeting inflammation pathways in the treatment of major depressive disorder. IL-6 could be a potential marker for patient profiling in personalized medicine approaches.