The level of ventilation (V̇E) at a given carbon dioxide output (V̇CO2) determines ventilatory efficiency. During cardiopulmonary exercise testing (CPET), ventilatory efficiency can be measured as the slope of the V̇E versus V̇CO2 relationship or the lowest V̇CO2. We evaluated the test–retest reliability of these two ventilatory efficiency indices in 29 healthy subjects (14 males). Each subject performed duplicate cycle ergometer tests on different days. Ventilation and the gas fractions for oxygen and CO2 were measured with a Vacumed metabolic cart. Linear regression analysis of the V̇CO2 slope for the duplicate tests in the males, females, and both sexes combined yielded correlation coefficients of 0·822, 0·942, and 0·910, respectively. The corresponding correlation coefficients for the lowest V̇CO2 were 0·745, 0·929, and 0·884. A comparison of the test–retest correlation coefficients between the two ventilatory efficiency measures for the men, women, and both sexes combined revealed that they were not significantly different and, for a given index, there were no sex differences. The bias (mean of difference scores between tests) and 95% limits of agreement for the V̇CO2 slope in the males, females, and both sexes combined were −0·05 ± 2·41, −0·57 ± 1·92, and −0·32 ± 2·20, respectively. The bias and 95% limits of agreement for the lowest V̇CO2 were very similar with values of 0·06 ± 2·45, −0·22 ± 2·03, and −0·10 ± 2·27. We conclude that the test–retest reliability for the V̇CO2 slope and the lowest V̇CO2 is the same and that there is no sex difference in reliability for either index of ventilatory efficiency.