Closing carbon mass balances is a critical and necessary step for verifying the performance of any conversion process. We developed a methodology for calculating carbon mass balance closures for a cellulase production process and then applied measurement uncertainty analysis to calculate 95% confidence limits to assess the accuracy of the results. Cellulase production experiments were conducted in 7-L fermentors using Trichoderma reesei grown on pure cellulose (Solka-floc), glucose, or lactose. All input and output carbon-containing streams were measured and carbon dioxide in the exhaust gas was quantified using a mass spectrometer. On Solka-floc, carbon mass balances ranged from 90 to 100% closure for the first 48 h but increased to 101 to 135% closure from 72 h to the end of the cultivation at 168 h. Carbon mass balance closures for soluble sugar substrates ranged from 92 to 127% over the entire course of the cultivations. The 95% confidence intervals (CIs) for carbon mass balance closure were typically +/-11 to 12 percentage points after 48 h of cultivation. Many of the carbon mass balance results did not bracket 100% closure within the 95% CIs. These results suggest that measurement problems with the experimental or analytical methods may exist. This work shows that uncertainty analysis can be a useful diagnostic tool for identifying measurement problems in complex biochemical systems.