There is little information on chemotherapy-induced peripheral neuropathy (PN) for community-dwelling patients with cancer. We studied 65,316 patients with breast cancer, 9242 with ovarian cancer, and 86,278 with non-small cell lung cancer from 1991 through 2002 identified from the 16 areas of Surveillance, Epidemiology and End Results. The incidence density of PN was 15.3, 21.5, and 18.3 per 1000 person-years for patients with breast, ovarian, and lung cancer who received platinum-taxane combination chemotherapy, respectively. Patients with breast, ovarian, and lung cancer receiving taxanes were more than twice as likely to develop PN compared with those not receiving chemotherapy (adjusted hazard ratio = 2.22, 95% confidence interval = 1.85-2.66 in patients with breast cancer), whereas patients who received platinum-taxane combination chemotherapy were more than 3 times as likely to develop PN compared with women who did not receive chemotherapy (adjusted hazard ratio = 3.33, 95% confidence interval = 2.05-5.05). In patients with ovarian or lung cancer receiving taxanes or platinum-taxane combination therapy, the risk of PN was increased with increasing number of chemotherapy cycles. These findings remained similar after adjusting for the history of preexisting PN or diabetes. Close monitoring for PN in patients receiving taxanes alone or in combination with platinum compounds may be warranted.