The analysis of metabolic fluxes of large stoichiometric systems is sensitive to measurement errors in metabolic uptake and production rates. It is therefore desirable to independently test the consistency of measurement data, which is possible if at least two elemental balances can be closed. For mammalian-cell culture, closing the C balance has been hampered by problems in measuring the carbon-dioxide production rate. Here, it is shown for various sets of measurement data that the C balance can be closed by applying a method to correct for the bicarbonate buffer in the culture medium. The measurement data are subsequently subject to measurement-error analysis on the basis of the C and N balances. It is shown at 90% reliability that no gross measurement errors are present, neither in the measured production- and consumption rates, nor in the estimated in- and outgoing metabolic rates of te subnetwork, that contains the glycolysis, the pentose-phosphate, and the glutaminolysis pathways.