Goodpasture disease (GP) is defined by the presence of anti-glomerular basement membrane (anti-GBM) antibodies and rapidly progressive glomerulonephritis. Besides anti-GBM, many patients with GP produce anti-neutrophil cytoplasmic antibodies (ANCA). For elucidation of the pathophysiologic significance of ANCA in this setting, epitope and antigen specificity of the anti-GBM antibodies and antigen specificity of ANCA were studied. Bovine testis alpha(IV)NC1 (tNC1); recombinant human alpha1, alpha3, alpha4, and alpha5(IV)NC1 (ralpha1 through ralpha5); and three chimeric proteins that contain previously defined epitope regions designated E(A), E(B), and S2 were used to examine the anti-GBM antibodies by ELISA in 205 Chinese patients with GP with or without ANCA. In the 205 anti-GBM antibody-positive sera, 63 (30.7%) were also ANCA positive (61 myeloperoxidase-ANCA and six proteinase 3-ANCA, four being triple positive). All 205 sera recognized tNC1 and ralpha3(IV)NC1. In the double-positive group, 54.0, 66.7, 71.4% of the sera could recognize ralpha1, ralpha4, and ralpha5, respectively, compared with 49.3, 60.6, and 55.6% for patients with anti-GBM antibodies alone. The levels of the antibodies to ralpha3, tNC1, and the alpha3/alpha1 ratio were lower in the double-positive group than that in patients with anti-GBM antibody alone (P < 0.05). Most of the sera could recognize the epitope regions E(A), E(B), and S2, but the absorbance values to E(A), E(B), and S2 were lower in double-positive group (P < 0.05). Double-positive patients had a broader spectrum of anti-GBM antibodies and lower levels of antibodies against alpha3(IV)NC1 compared with that of patients with anti-GBM antibodies alone.