The basin of the Liwiec River is one of archaeologically least investigated areas of central-eastern Poland (between the Vistula, Bug and Wieprz Rivers). Evidence from the 12th and 13th c. written sources is lacking. Regular study of the Early Medieval pattern of settlement on the Liwiec was initiated only in 1983 when excavation was started of the Early Medieval earthwork at the village of Grodzisk (Phase 1 – 10th c., phase 2 – mid 11th c. to the beginning of 13th c.). In the present study an attempt is made to define the relative chronology of Early Medieval earthworks on the Middle and Lower Liwiec with the help of data obtained from the investigation study of the region’s geology and geomorphology rather than strictly archaeological evidence. A major challenge in the interpretation of archaeological evidence recovered from the Liwiec cachment is how to accurately define its chronology. Over 90% of finds dated to the Early Medieval period is made up by pottery which unfortunately is not a very sensitive chronological indicator. Moreover, pottery finds originate from outside closed assemblages making it difficult at present to elaborate a clear scheme of relative chronology. The generally used system of dating Early Medieval pottery within the framework of absolute chronology has a margin of error of 100–200 years. This poses a serious problem when studying the processes of settlement dynamics. There is a need therefore to seek other methods for defining chronological phases of local 10th–13th c. settlement network in the Liwiec cachment. In investigating the geology and geomorphology in the Liwiec valley the aim was to determine the water levels in the river in the Early Medieval Period. Detailed analysis was made in the middle and lower course of Liwiec, which have well defined flood terraces. These are useful in reconstructing the natural history of the river during the last several hundred years. With a length of 126 km and cachment of ca. 2.770 km² Liwiec is a swiftly flowing stream registering substantial (up to 3 m in its lower course) fluctuation in its water levels and strongly meandering as well. Liwiec has developed numerous old channels, especially in its middle and lower course, branches and stretches of standing water. In its upper stretch the valley is much less well-defined as a result of hydrographic conditions changing over time. The first series of geomorphology studies in the Liwiec cachment was carried out in 1987 at the Early Medieval earthwork in the village of Grodzisk. It was established that in the period when the earthwork was in use he channel of the Liwiec was 6–7 m higher than at present. Gradual erosion of the river bed started at the turn of the 12th and the 13th c. leading to a marked drop in the water level in the river. All of this suggests that the 14th c. castle at Liw cannot have had a predecessor in the form of an Early Medieval – 12th c. earthwork. In the period when the Grodzisk earthwork was still in function (12th/13th c.) the site subsequently occupied by the castle at Liw, which lies within the valley would have been below the waterline. The next step was a study of the geology of other four Early Medieval earthworks situated on the Middle and Lower Liwiec: Barchów (previously dated to the 10th–11th c.), Liw (12th c.?), Grodzisk and Wyłazy (11th––12th c.). Except for Grodzisk all these sites had been investigated to a limited extent by sondage and were dated primarily on the basis of the modest quantity of pottery finds. The study included also three other conjectured no longer existent Early Medieval earthworks (Łochów, Węgrów and Kucyk, previous localisation: Wyszków) which tend to be overlooked by researchers. Their approximate location in the field was established following a careful analysis of meagre late 19th and 20th c. references. Researchers concerned with Early Medieval settlement in Poland generally assume that in the older phases of this period (6th–9th c.) the groundwater table and water levels in rivers were considerably lower than today and that this made it possible for villages and earthworks to be set up in river valleys. According to this concept settlement started to move out of the valleys towards the valley margin with the rise in water levels, i.e., from the 10th c. onwards. Nevertheless, a number of 11th–13th c. settlements is known which are found on the river valley floor. It may be added that geologists and hydrologists have yet to reach a consensus on the subject of groundwater levels and development of river valleys in the youngest stages of the Holocene. Consequently, the results of the geomorphologic study presented here may be related only to the analysed stretch of the Liwiec Valley and should not be mechanically be transposed onto other areas. The second series of investigations of geology and geomorphology was carried out in 1989. Its results fully confirmed the conclusions reached in 1987 with regard to Grodzisk. Only the earthwork at this site lies outside the river valley on the edge of the upland. A similar location, on the highest, fourth, terrace is seen at Wyłazy. The establishment at Barchów and the castle at Liw occupy a younger, third river terrace. The oldest flood terrace in the valley of the Liwiec is the at present poorly preserved terrace IV. Terrace III is slightly younger while terraces I and II are much later developments. Assuming the gradual decline of the water levels in the Liwiec it may be supposed that the oldest of the studied earthworks is the one at Grodzisk, rising on the edge of a marshy and river valley unsuitable for settlement. Lowering of the water level in the river and development of terrace IV made settlement possible also in the valley of the Liwiec; consequently, the earthwork at Wyłazy would have developed presumably still in the period when the Grodzisk earthwork was in function. Decidedly the youngest earthwork would be Barchów, situated on terrace III while Liw is an establishment dating to the mature Medieval Period (14th c. onwards). Analysis of the geology and geomorphology of earthworks in question helped to establish their relative chronology. In the proposed scheme, earthworks at Grodzisk, Wyłazy and Łochów are dated to the oldest horizon I, while earthworks at Kucyk, Barchów and Węgrów(?) date to the younger horizon II. Both chronological horizons fit in the period between the 10th and the 13th c. In horizon II there was a visible concentration of stronghold settlement in the middle course of the Liwiec; perhaps functions served by the earthwork at Wyłazy were taken over after it was deserted by the earthwork at Podnieśno, previously dated to the 11th c. For the studied stretch of the Liwiec it may be assumed that the process of colonisation of the river valley proper took place parallel with the falling water levels only in the younger phases of the early Middle Ages. Both the analysis of archaeological data and of indirect written evidence relating to administrative and church territorial divisions of this part of Poland in the early Middle Ages suggest that the area lying between the Bug and Liwiec had a frontier, Polish-Ruthenian character. Data available is too imprecise to attempt reconstruction of the Polish-Ruthenian frontier in the region. But we may assume that most probably from the middle of the 11th c. a frontier belt started taking shape here, open in character, following the line of the water divide of both the rivers in question. The relative chronology of early Medieval earthworks lying on the Middle and Lower Liwiec presented here is a further step leading to a more precise than at present definition of the character and course of the Polish-Ruthenian border area between the 11th and the 13th c. in the region.