OBJECTIVES: A randomized trial comparing doxetaxel (D) and paclitaxel (P) treated patients has shown greater neurologic toxicity with P given weekly.  This study aimed to compare neuropathy (NP) and NP-related outcomes between D and P in a community practice. METHODS: The analysis was conducted using the Georgia Cancer Specialist Database (2003-2008) supplemented by chart abstraction. Patients with stage I-III breast cancer (BC) and treated with adjuvant D or P-containing chemotherapy (CT) were followed from initial treatment to the earliest of death, loss to follow-up, or switch to other regimens.  NP and NP-related outcomes were measured as (1) 6-month rate of severe NP (CTC grade 3/4), (2) 6-month rate of receiving NP-related medications (NPMs), (3) proportion of patients whose CT was affected due to NP (CT being held, discontinued, or dose reduction). Univariate and multivariate analyses with adjustment of clinical and demographic characteristics were performed to compare the outcomes between D and P groups.  RESULTS: A total of 591 (63.6%) and 338 patients treated with D and P were included. The univariate analyses showed that P had higher rates of severe NP (4.9% vs. 1.0%) and NPMs use (28.1% vs. 7.6%) within 6 months, and their CT was more likely to be affected (7.1% vs. 1.5%), compared with D group (all p<0.001).  Similarly, the adjusted results (all p<0.001) showed that P was at higher risks for severe NP (HR=3.50), NPMs use (HR=6.40), and CT interruption (OR=5.40). CONCLUSIONS: This study found that P was associated with higher risks for severe NP and CT interruption, compared to D. The findings are consistent with published trials and call for close observation of NP when using P. The results should also be viewed in the context that NP poses higher economic burden over non-NP patients. The findings need to be confirmed by addition studies from community settings.