Introduction: Febrile neutropenia (FN) is a major dose-limiting toxicity of myelosuppressive chemotherapy that can result in hospitalization, dose reductions or treatment delays and compromised clinical outcomes. National Comprehensive Cancer Network guidelines recognize that the risk of developing FN is based on a regimen’s risk (high, intermediate, low) and patient-related risk factors. This study aims to estimate the association of demographics, comorbidities, chemotherapy regimen and FN risk factors on the risk of developing FN in the first cycle among a cohort receiving pegfilgrastim prophylaxis, for the purpose of informing real-world comparative effectiveness studies. Method: Patients with breast cancer or non-Hodgkin lymphoma (with CHOP±R) receiving prophylactic pegfilgrastim from 2017/1/1 to 2018/5/31 were identified from MarketScan ® Commercial and Medicare databases. Patients were required to be continuously enrolled 6-months prior to and post the index pegfilgrastim prophylaxis date. FN was defined using claims with a diagnosis code for neutropenia (main definition), neutropenia with fever or infection (sensitive definition) and neutropenia with fever (specific definition). FN incidence among the first chemotherapy cycle were estimated using logistic regression after controlling for cancer type, region, insurance type, regimen level for FN, Quan-Charlson comorbidity score, number of days from chemotherapy start to pegfilgrastim prophylaxis and number of FN risk factors. Results: A total of 4,807 patients (96.9% breast cancer) were identified. Mean age was 53.5 years old and 95.8% were female. Most of the patients were treated with regimens with a high risk of FN (78.5%) and had at least one risk factor (71.4%). The odds ratio of FN for patients with 3 vs. 0 risk factors were 5.4, 5.6 and 6.5 for main, sensitive and specific definitions, respectively (all p <0.01). Patients treated with intermediate-risk regimen were more likely to experience FN than high-risk regimen (odds ratio=1.6, 1.7, 1.8 for main, sensitive and specific definitions, all p <0.05). Conclusion: Despite prophylaxis with pegfilgrastim, patients with multiple risk factors and those treated with chemotherapy regimens at intermediate risk of FN are still at risk of developing FN. Real-world studies evaluating the comparative effectiveness of myeloid growth factors must apply statistical methods to mitigate bias by considering these variables when adjusting the cohorts.