We have expressed conformationally intact, enzymatically active recombinant PR3 in HMC-1 cells (HMC-1/PR3 cells) that is recognized by C-ANCA. Here we directly compared the clinical utility of C-ANCA testing by indirect immunofluorescence (IIF) using HMC-1/PR3 cell cytospin versus polymorphonuclear neutrophil (PMN) cytospin preparations and commercially available anti-PR3 ELISA kits. Two hundred sera were tested independently by three investigators: 101 previously determined to be C-ANCA-positive by routine clinical laboratory testing using standard IIF on PMN cytospins, and 99 control samples chosen primarily because they contained antibodies against other cytoplasmic target antigens. Discrepant test results between the two cellular substrates were found in seven samples: 2/7 were PMN-positive and HMC-1/PR3 cell-negative (one Sjogren's syndrome, one hand injury); 5/7 were PMN-negative and HMC-1/PR3-positive (all Wegener's granulomatosis (WG)). All C-ANCA-positive WG patients were also positive on HMC-1/PR3 cells. IIF using HMC-1/PR3 cells was as sensitive as the most sensitive anti-PR3 ELISA (79.8% versus 80.7%, P = 0.739), and more sensitive than standard IIF C-ANCA testing using PMN cytospins (79.8% versus 75.2%, P = 0.025) or the anti-PR3 ELISA with the least false-positive test results (79.8% versus 63%, P < 0.01). These findings indicate that HMC-1/PR3 cells are a very sensitive antigen-specific substrate for clinical anti-PR3 ANCA testing which appears superior to standard C-ANCA testing using PMN cytospin substrates and anti-PR3 ELISA. Our results also suggest that in WG the C-ANCA fluorescence pattern is not caused by antibodies against target antigens other than PR3.