Two previously independent treatments of length-temperature time normalization in gas chromatography have reached different conclusions as to the magnitude of the optimizing capacity ratio, one yielding the value of 2 and the other value of 3. This paper deals with the inconsistency. It is demonstrated that whether one obtains K′opt = 2 or 3 depends upon the assumption one makes regarding the column permeability and the dependence of the HETP on the carrier velocity. These two values are the two possible extremes. Real systems, it is shown, have an optimum capacity ratio between those two values, depending upon the experimental conditions. It is also shown that in studying plate height dependence on the column length, especially in capillary columns, the usage of constant average flow rather than constant outlet velocity can be misleading. Finally, it is indicated that as k′ changes from 1.5 to about 5, the analysis time varies only by about 10%.