Phase 1, a circular earthwork monument, constructed around 2950-2900 BC, comprised a ditch with an internal bank defining an area about 90m across. Immediately inside the bank was a ring of 56 equally-spaced holes (the Aubrey Holes) believed to have contained upright posts. Outside the ditch was a small counterscarp bank. There were at least three entrances. Deposits of animal bones were placed on the bottom of the ditch in some areas, with particular emphasis on the entrances. An organic dark layer formed over the primary silting of the ditch (Cleal et al. 1995,63). It may be noted that the construction of the Phase 1 enclosure at Stonehenge is broadly contemporary with the construction of the bank and ditch at Avebury (Pitts and Whittle 1992,205).
Stonehenge World Heritage Site - Archaeological Research Framework Version 2 – April 2002 Section 2– Page 11 of 59

Phase 2, 2900-2400 BC, the basic structure remained the same, but there is evidence for the deliberate back-filling of parts of the ditch, natural infilling, and some features cut into the fills. The Aubrey Holes survived as partly filled features lacking posts by this stage, but timber settings were constructed in the centre of the monument, at the north-eastern entrance, near the southern entrance and outside the earthwork boundary to the northeast. Towards the end of the phase, cremation burials were deposited in the Aubrey holes, the upper ditch and around the circumference of the monument, on and just within the bank (Cleal et al. 1995, 115).

Phase 3i, broadly 2550-2200 BC, the first stone phase of the monument is built with the erection of a setting of paired bluestones, the plan of which is far from certain, in the Q and R holes roughly in the centre of the space defined by the earlier earthwork. The main entrance to this structure was to the northeast and was marked by additional bluestones set inside the double
circuit. It is possible that a large slab of greenish sandstone, the Altar Stone, was the focus of this structure; other stones may have stood within and around it. External to the earthwork enclosure it is likely that at least the first straight length of the Avenue belongs to this phase, as too stone settings around the entrance.

Phase 3ii-v, broadly 2400-2000 BC, sees the demolition of the Phase 3i structures and their replacement (perhaps gradual) by an arrangement of four concentric stone settings which from the inside working outwards comprise: the bluestone horseshoe, five sarsen
trilithons arranged in a horseshoe, the bluestone circle, and the sarsen circle. This is the stone structure that can be seen in a ruined state today. Modifications were also made to the peripheral arrangement of stones and perhaps the Avenue. The burial of an adult male with evidence of traumatic pathology suggesting death caused by arrow-shot was found in a grave dug into the ditch of the northwest sector dates to about 2400-2140 BC (Evans 1984; Cleal et al. 1995, 533). The apparent integrity of the phasing of Stonehenge and its associated structures hides a great deal of uncertainty. Only a small proportion of the features has been dated, and some key events have very few associated dates. The distribution of elements over a large area limits the use of horizontal and vertical stratigraphy. The longevity of the sequence at the site inevitably introduces problems of residuality in the disposition of finds and datable material. Especially difficult issues include the relationship between the Phase 1 and Phase 2 features; the form and plan of the Phase 3i structure; the sequence of construction for individual elements of the Phase 3ii-3v settings; the sequence and arrangement of features in the centre of the monument (cf. Burl 1997; 2001); the sequence and arrangement of stone settings within and around the northeast entrance (cf. Burl 1991; 1994; Pitts 1982); and both the internal phasing of the Avenue construction and the links between these and the development of the stone settings (cf. Cleal et al. 1995, 533-4).

Courtesy of Tony Johnson (Books.php)