The stratospheric major warming of early December 1987 is analyzed using NMC observations of temperature and geopotential. This warming is distinguished as the earliest major warming ever recorded in the Northern Hemisphere winter. The observed mean zonal wind reversal and reversed poleward temperature gradient at 10 mb were preceded by the anomalous amplification of a zonal wavenumber 1 planetary wave emanating from the troposphere. This planetary wave event, similarly, is distinguished for having produced the second largest sustained flux of wavenumber 1 activity ever observed to propagate upwards from the troposphere at such an early time in the winter. Within the troposphere, the amplification of wave 1 was accompanied by several simultaneous blocking episodes, but it is unclear whether this blocking caused the anomalous formation of the planetary wave (or vice versa, or neither). Amplification of the planetary wave within the stratosphere led to a significant off-polar displacement of the circumpolar vortex and a reduction in vortex area, as observed in connection with other warmings. However, in the present case there is less significant evidence of a preconditioned stratospheric vortex, except for a small precursor event in late November which may have slightly retarded the normal, climatological expansion of the vortex. Therefore, it appears that this unusually early major warming was mainly attributable to an anomalously large tropospheric forcing. After the climax of the warming, the midstratosphere vortex was observed to split into a double-vortex pattern. This feature is quite striking when viewed three-dimensionally, as the two 850 K vortex components remained contiguous, respectively, with single vortices in the upper and lower stratosphere. Thereafter, the stratosphere returned to a cold, undisturbed pattern until the beginning of March, when an early final warming occurred. The relatively cold January-February period coincided with the deep westerly phase of the equatorial quasi-biennial oscillation (QBO), as observed in connection with other cold, undisturbed winters. However, the QBO had already attained this phase by early December 1987, suggesting that the phase of the QBO per se is insufficient to prevent the occurrence of a major warming.