Purpose The purpose of this study is to (a) visualize the symptom–cytokine networks (perceived stress, fatigue, loneliness, perceived cognitive impairment, daytime sleepiness, sleep quality, and 13 cytokines) and (b) explore centrality metrics of symptom–cytokine networks in breast cancer survivors who completed chemotherapy treatment. Methods Cross-sectional analysis of data collected from 66 breast cancer survivors who were on average three years post chemotherapy completion. Perceived stress, fatigue, loneliness, perceived cognitive impairment, daytime sleepiness, and sleep quality were measured with self-report instruments, and a panel of 13 cytokines was measured from serum using multiplex assays. Symptoms and cytokines were simultaneously evaluated with correlations, network analysis, and community analysis. Results Network analysis revealed the nodes with the greatest degree and closeness were interleukin-2, granulocyte-macrophage colony-stimulating factor, interleukin-13, and perceived cognitive impairment. Node betweenness was highest for perceived cognitive impairment and interleukin-2. Community analysis revealed two separate communities of nodes within the network (symptoms and the cytokines). Several edges connected the two communities including perceived cognitive impairment, stress, fatigue, depression, interleukin-2, granulocyte-macrophage colony-stimulating factor, interleukin-8, interleukin-13, and interleukin-10. Partial correlation analyses revealed significant negative relationships between interleukin-2 and fatigue, loneliness, stress, and perceived cognitive impairment ( rs = −.27 to −.37, ps < .05) and a significant negative relationship between perceived cognitive impairment and granulocyte-macrophage colony-stimulating factor ( r = −.34, p < .01). Conclusions Our analyses support that perceived cognitive impairment, stress, loneliness, depressive symptoms, and fatigue co-occur and extend the literature by suggesting that interleukin-2 may contribute to the underlying mechanistic pathway of these co-occurring symptoms. Our findings add to a growing body of literature that is shifting to study symptoms as they co-occur, or cluster, rather than individual symptoms.