We compared the standard nitrogen (N2) washout technique for measuring functional residual capacity (FRC) with a modified technique that uses a helium/oxygen mixture (heliox) at different ratios instead of pure oxygen. The tests were made with a standard lung function system equipped with an ultraviolet (UV) analyzer for measurement of N2 concentrations in the expired gas. We examined models of “spontaneous breathing” and “mechanical ventilation,” each with volumes of FRC in the range of a premature and a newborn lung (20–80 ml), using both techniques at different baseline inspired oxygen concentrations (FiO2). Correlations between known and measured volumes were high and identical for the two techniques (r = 0.996), and the mean error was not significantly different from zero (P = 0.111). Measurements of FRC in 6 infants gave a correlation coefficient of r = 0.989 between the two techniques; reproducibility, as measured by the coefficient of variation, was high, showing no significant differences between both techniques (P = 0.792). However, values of individual infants were different (P = 0.011), and the slope of the regression line relating measurements by the 2 techniques was 1.04, with an intercept on the y‐axis at 1.46. We conclude that FRC can be measured with the modified N2 washout technique, using heliox as a washout gas. Volumes can be measured with high precision and reproducibility, even in premature infants with low lung volumes and/or high baseline FiO2. A correction factor may be necessary to equate FRC measurements made by oxygen‐N2 vs. heliox‐N2 washouts. Hyperoxemia and hypoxemia can be avoided by admixing different flows of oxygen to a standard heliox mixture. Pediatr. Pulmonol. 1997; 23:434–441. © 1997 Wiley‐Liss, Inc.