The variability of the energy cycle in the atmosphere is studied for various time scales, namely, year-to-year, season-to-season, and month-to-month, by using daily aerological data for the 5-year period, May 1958 through April 1963. An objective method of analysis is used throughout the study. Our earlier analyses of the normal annual cycle of the energetics are expanded with seasonal and monthly estimates for the different years. It is found that the local values of the integrands of the hemispheric mean eddy available potential energy and of the hemispheric conversion from mean to eddy potential energy are positively correlated. Similarly, the local values of the integrands of the hemispheric mean eddy kinetic energy and of the hemispheric conversion from eddy to mean kinetic energy are positively, although not very highly, correlated. In the case of both potential and kinetic energy, most of the activity is found in preferred geographical areas, namely, in middle latitudes east of the major continents. The amplitude of the interannual variations in the monthly hemispheric integrals of all four basic energy components is generally less than 10% of the normal annual amplitude. The interannual variability in the various forms of energy and in the corresponding energy transformations is largest in winter and smallest in summer. The hemispheric eddy potential and eddy kinetic energy are found to be very highly, positively correlated. Considering the deviations from the normal annual variation for the hemispheric mean eddy potential energy and for the hemispheric conversion from mean to eddy potential energy, they are found to be positively correlated with each other. On the other hand, the corresponding deviations for the hemispheric eddy kinetic energy and for the hemispheric conversion from eddy to mean kinetic energy seem to be negatively correlated. In general, the monthly energy conversion integrals show a much higher year-to-year variability than the integrated energy amounts. Since on a monthly basis the anomalies in the different conversion terms do not correlate well with each other, at least part of this large year-to-year variability must be spurious and due to uncertainties in the initial data and/or the later processing of the data. Because for the time scales considered it seems difficult to follow the sequence of events in the energy cycle, and thereby to unravel cause and effects, clarification of these questions will have to wait until reliable daily analyses of the full energetics will be possible.