Objective Cytokines may be linked to depression, although it has been challenging to demonstrate this association in cancer because of the overlap between depressive symptoms and other sickness behaviors. This study investigates the relationship between cytokines and depression in cancer patients, accounting for confounding clinical and methodological factors. Methods The GRID Hamilton Rating Scale for Depression and Neurotoxicity Rating Scale (NRS) for cytokine‐induced sickness behaviors were administered to 61 cancer patients and 38 healthy controls. The cancer group was of mixed type and largely of late stage, with a recruitment rate of 35% and completion rate of 47%. Major depression was diagnosed in 19 of 61 (31%) cancer patients. Multiplexed cytokine assays for inflammatory and anti‐inflammatory cytokines were conducted in plasma samples using electrochemiluminescence. Results All cancer patients had high NRS scores and elevated levels of most cytokines. Cancer patients with major depression had higher NRS scores than those without major depression. IL‐1rα was positively associated with the GRID scores of depressive symptoms (regression coefficient, 3.52 ± 1.18; P = .004), but not with major depression. Major depression was negatively associated with the anti‐inflammatory cytokine IL‐4 (regression coefficient, −0.65 ± 0.26; P = .013), but not with IL‐1rα. Conclusions Depressive symptoms in cancer patients may represent sickness behaviors, which may have distinct cytokine associations from major depression. Sickness behaviors may be associated with an increase in inflammatory cytokines, whereas major depression may be induced by a failure to adequately resolve inflammation. Our findings suggest that cytokine‐mediated interventions may be of value to treat depression in this population.