The accuracy of the formula derived by Drorbaugh and Fenn (Pediatrics 16: 81-86, 1955) for calculating tidal volume (VT) from the phasic pressure change measured when an animal breathes in a closed chamber has recently been challenged. Epstein and Epstein (Respir. Physiol. 32: 105-120, 1978) argue that the formula may underestimate VT by up to 30% and predict that the error increases as the ratio of inspiratory duration (TI) to total breath duration (Ttot) increases, and as the expired temperature at the nares (TN) increases. To test their theory, I measured VT in anesthetized rats by the barometric technique and by conventional pneumotachography simultaneously. TN was varied from ambient to body temperature by passing a variable current through the pneumotachograph heater; TI/Ttot was varied by changing FICO2 and by selecting different rats. The predictions were confirmed. A factor is derived for retrospectively correcting VT estimated by the Drorbaugh-Fenn formula. It requires knowledge of TN and TI/Ttot and reduces the error between experiments to under 20% and within each experiment to about 5%. To facilitate its use, TN was measured in rat, rabbit, cat, man, and infant pigtail monkey.