We compared troponin I (TnI) assays (AxSYM [Abbott]; ACS:180 [Bayer]) in blood samples with concentrations less than 10 ng/mL (< 10 micrograms/L). Discordant results were evaluated by linearity studies and by testing for rheumatoid factor. Patients with discordant TnI results were compared with patients with concordant results and patients with negative TnI who had a new myocardial infarction or died within 2 months of initial testing. Positive TnI cutoffs by AxSYM and ACS:180 were 0.7 ng/mL (0.7 microgram/L) and 0.13 ng/mL (0.13 microgram/L), respectively. We identified 173 specimens that were repeatedly positive by at least 1 assay; 143 specimens were positive by both assays. Twenty samples positive for TnI by AxSYM were negative by ACS:180, while 10 samples positive by ACS:180 were negative by AxSYM. The discordant samples showed no evidence of interfering substances, including rheumatoid factor. Clinical follow-up showed that 26% of patients with elevated TnI by both assays, 33% with TnI positive only by AxSYM, 22% with TnI positive only by ACS:180, and 8% with negative TnI by AxSYM encountered at least 1 clinical end point. Variable detection rates by these assays for low-positive TnI represent a clinically significant problem.