Using reported height data at mandatory pressure levels, a version of Hoskins' Q-vector analysis is computed for two layers, producing contoured patterns of vertical motion and other diagnostics at two levels. This analysis, based on quasi-geostrophic theory, is applied to a severe convective storm episode (25–26 June 1982) in eastern Colorado that resulted in the development of a mesoscale convective complex. Convection developed in association with a polar air mass overlying the high plains and a weak short wave in the westerlies. LFM/MOS and quantitative precipitation guidance indicated eastern Wyoming and western Nebraska as the most threatened area, whereas the actual convection was most intense in Colorado, and the resulting convective complex propagated into Kansas, weakening as it did so. Another convective system developed in northeastern New Mexico, later propagated southeastward into the Texas panhandle and intensified. Q-vector diagnostics correctly indicated similar changes in the large-scale forcing at least 4 h before the observed changes in the convective systems occurred, whereas LFM-predicted vertical motions indicated the Kansas storms would dominate. When viewed in light of concurrent information such as satellite images and surface maps, the Q-vector diagnostics suggest how NMC and NSSFC guidance could have been modified in this case. Omega diagnostic patterns computed from 3-hourly soundings (A VE-SESAME V, 20 May 1979) indicate for the most part reasonable temporal continuity over a period of 6 to 12 h, thus lending some justification to their potential use as a forecasting aid. Further experience and development are necessary to determine the scheme's value under a variety of synoptic conditions. However, the fact that it currently runs on a microcomputer in only 13 min indicates that it may have potential as a real-time operational diagnostic aid for short range forecasting.