The Gluhite kamani Thracian rock -cut complex in the Eastern Rhodopes Mountain is famous with its more than 500 rock niches carved in the rock at heights up to 20 m. Continuous habitation starting from the final Chalcolithic (the second quarter of the 4 th millenium BCE) until the Late Classical - Early Hellenistic period (late 5 th - 4 th century BCE) provides unique opportunity for studying diachronic changes in pottery production and use. In this contribution, systematic analysis of a collection of 135 ceramic sherds from five archaeological periods at the site was carried out using a combined mineral magnetic and geochemical approach. Magnetic susceptibility (chi), frequency -dependent magnetic susceptibility (chi fd %) and firing temperature (T fire ), determined using the magnetic susceptibility method, were combined with data from elemental XRF analyses. Thermomagnetic measurements, employed for identification of magnetic minerals revealed the dominant presence of (titano) magnetite with low Ti -content and/or a fraction of single-domain magnetite and in few cases hematite. Although varying in a wide range, median chi values of ceramic sherds from the Early Iron Age (EIA) periods show similar distribution, while for Late Iron Age (LIA) with the introduction of the wheel-thrown pottery, chi values are higher, presumably due to decreased amount of silica-rich clastics in the clay. Concurrently, chi fd % values decrease in ceramics from the LIA, indicating predominant presence of stable single-domain magnetite in the iron oxide composition of the ceramics. Firing temperatures (T fire ) estimations display a systematic pattern of two groups - higher (T fire between 760 and 800 degrees C) and lower (T fire - 400---600 degrees C). During the EIA periods both T fire groups are present, characterized by systematic gradual decrease of the median value of the low T fire when going from EIA1 towards the EIA2. Wheel-thrown pottery from the LIA show low T fire - 400 degrees C only. Based on the differences in thermomagnetic properties of ceramics, their magnetic parameters and T fire estimates, two possible hypotheses are suggested to explain the observed changes in EIA ceramic characteristics: i) utilization of two differently sourced clay materials (illite-dominated and montmorillonite-dominated) for production of ceramics fired at higher and lower temperatures, respectively; and ii) utilization of one raw clay source, while using two firing regimes - firing at higher temperature for prolonged time, and fast firing to high temperature with successive stage of longer firing at lower temperature. Differences in thermomagnetic properties of wheel-thrown pottery from the LIA period leads to the assumption for utilization of diverse clay source for their production, as compared clay sources for the Early Iron Age periods.