Myeloid cells are immune cells that play a complicated double role in cancer — sometimes helping tumors grow and sometimes fighting them. Scientists have long assumed this was because different subtypes of myeloid cells exist within tumors. This review challenges that assumption, arguing that 'plasticity' — the ability of individual myeloid cells to shift their behavior — is actually the more fundamental driver. The authors identify two forms of this plasticity: one involving how myeloid cells develop (differentiation plasticity) and one involving how they function once mature (functional plasticity). Understanding these mechanisms could help explain why many current cancer immunotherapies fail, and opens new avenues for treatments that reprogram myeloid cells rather than simply trying to block or deplete them.