Apart from Mureybet, Abu Hureyra is one of the few sites on the middle Euphrates where Epipalaeolithic and early Neolithic layers have been found. The radiocarbon dates of the Epipalaeolithic layers (Fig. 1) correlate quite well with the stratigraphy. Only OxA 475 of the uppermost layer E 396 and BM 1719R seem to belong to a transitional phase to the late PPNB settlement.The range of the radiocarbon dates of the Epipalaeolithic layers indicates an occupation of the settlement between 11220 and 10750 BC.
All radiocarbon dated domesticated cereals (red oval) that were found in the Epipalaeolithic layers seem to be intrusive from upper layers. They were all dated to a time range between 7450-7070 BC. The only exceptions are the “domesticated-like” rye seeds considered by Hillman to be fully domesticated (Hillman 2000; cf. Weiss et al. 2006). It is striking that no other cereal except rye seems to have been domesticated during the Epipalaeolithic. The uncalibrated radiocarbon dates of the Neolithic settlement of Abu Hureyra (Fig.2) cluster into two phases: samples dating to the older phase were predominantly found in Trenches B and D, and those from the younger phase from Trenches E and G, as was noticed by Moore (2000:257). 
Fig. 2 also shows that Trench C likewise dates from Phase 2A. Even the radiocarbon date BM 1425 of Trench C does not overlap after calibration with all the later dates of the second phase, which overlap in the 1-σ range up to OxA 6418: 8115+/-80. The obviously recent date OxA 877: 8300+/-150 BP of Trench D, which lays in the range of the later dates of Trench E and G before calibration, overlaps after calibration in the
1-σ range with the older dates of Trench D.
Up to April 2009 only one date from Trench A had been published. It clearly is from phase 2B.
This shift from Trenches B, C and D to Trenches A, E and G is also visible after calibration, though due to the probability ranges of the calibrated radiocarbon dates the separation of the dates becomes less clear.
The main occupation of Phase 2A probably starts at about 7760 BC or, as two AMS-dates of cereals show, even later, at about 7600 BC. Phase 2A lasts until around 7440 BC (cf. Moore 2000:256).
The second concentration of dates starts around 7400 BC and lasts until 7100 BC. Although most of the radiocarbon-dated cereals from Phase 2B were found in the Epipalaeolithic layers, they were certainly displaced from upper Neolithic layers. They point to a clearly distinguishable Neolithic phase and correlate well with other in situ samples of this phase.
It is thus possible to include these “aberrant” samples in the dating of Phase 2B. Whereas the charcoal samples scatter over a wide range, the cereal samples cluster in a narrow time frame. The latest date (OxA 1931: 7890+/-90 BP) comes from Level 2 of Trench G, but the sample is probably displaced. Like the other intrusive cereal samples of the Epipalaeolithic layers, it further illustrates the extensive dislocations due to pits and buildings dug from later into earlier layers.
To sum up, two Neolithic phases can be distinguished that document a shift of the used space within the settlement. According to the radiocarbon dates, it can be postulated that settlement activities shifted from Trenches B, D, and C (Phase 2A) to Trench E, G, and A (Phase 2B). There is only one date (BM 1424) of Phase 2A Trench B that points to a use during the later Phase 2B.
Due to the large time ranges, the radiocarbon dates function like a scrim diffuser and suggest a “continuous” occupation. However it is not possible to determine whether the settlement was permanently occupied or whether there were instead several short occupation phases interrupted by brief periods of abandonment.
The separation of two absolute chronological Neolithic phases is supported by the observation of Moore et al. (2000:257), who distinguished Phase 2B from Phase 2A essentially because of the dominance of sheep and goat in Phase 2B, in contrast to Phase 2A, during which gazelle were still hunted.
It is difficult to determine the date of the abandonment of the settlement. Most of the younger dates of the Neolithic Phase 2B end around 7100 BC; but this is probably due to the course of the calibration curve, which is quite flat until 7100 BC and than drops steeply. The occupation might therefore end earlier. The short-lived cereal samples suggest two relatively brief Neolithic occupation phases, but it has yet to be verified that these two phases are also reflected in the anthropological data.
BM 1719R and OxA 475 were most probably dislocated from more recent layers. BM 1719R also might be considered an example of “old wood effect”. (Mithen and Finlayson 2007 discussed the problem of dating old wood for Wadi Faynan s.p. 31). Two other samples (BM 1122 and OxA 1288) of the Neolithic layers also date to this transitional phase.
 The different phases in the diagram (Fig.2) do not correspond to the phases of Moore et al. (2000:257). In order to compare the layers of different trenches, it was necessary to create a new system relying on the stratigraphic positions of the samples.
 Two dates, BM 1724R and OxA 1931, are much younger than all the other dates and probably point to a sporadic use of the site after 7000 BC.
 The sum of the dates of Phase 2A was calculated without the sample OxA 877, as it could not be attributed clearly to either Phase 2A or 2B. The ranges of the sum for Phase 2A are: 7790-7770 BC (0.8%); 7760-7440 BC (63.0%); 7410-7360 BC(4.4%). For Phase 2B they are: 7450-7400 BC (7.9%); 7370-7120 BC (55.2%), 7110-7080 BC (5.1%). The charcoal sample of trench B, BM 1122: 9374 +/- 72 BP dating to around 8500 BC, is probably too old, but it could also derive from the so- called “intermediate period” indicated by several other dates (cf. footnote 2).
Fig. 2 Uncalibrated radiocarbon dates of Neolithic phases from Abu Hureyra, arranged according to tenches and material. "Phase" indicates their stratigraphic position within the trenches (legend s. Fig. 1).