Background: Given the limited literature, we conducted a study to examine the epidemiology of herpes zoster (HZ) among newly diagnosed cancer patients.Methods: We identified adult health plan members of Kaiser Permanente Northern California diagnosed with invasive cancer from 2001 to 2005. Electronic health records with inpatient and outpatient diagnoses, laboratory tests, and antiviral medications were used to identify HZ diagnoses from 2001 to 2006. HZ diagnoses and associated complications were confirmed by medical chart review. Treatment with chemotherapy and corticosteroids was used to classify patients by immunosuppression level.Results: Among 14,670 cancer patients, 424 were diagnosed with HZ during follow-up (median 22 months). The incidence of HZ was 31/1,000 person-year (PY) in patients with hematologic malignancies and 12/1,000 PY in patients with solid tumors. The corresponding 2-year cumulative incidence of HZ was approximately 6% and 2%, respectively. Compared with incidence rates of HZ reported in a general US population, the age-and sex-standardized rates of HZ were 4.8 times higher [95% confidence interval (CI), 4.0-5.6] in patients with hematologic malignancies and 1.9 times higher (95% CI, 1.7-2.1) in those with solid tumors. HZ risk increased with increasing level of immunosuppression. Among HZ cases, 19% with hematologic malignancies and 14% with solid tumors had HZ-associated pain for at least 30 days. The corresponding numbers for nonpain-related complications were 30% and 18%, respectively.Conclusions: Cancer patients are at substantially increased risk of HZ and among those with HZ, complications are relatively common.Impact: Better HZ prevention and treatment options for cancer patients are needed. Cancer Epidemiol Biomarkers Prev; 22(1); 82-90. Ó2012 AACR.