Abstract Archaeological stratigraphies that may contain floors are often encountered. For more than a decade soil micromorphologists working on archaeological sequences have benefited from having access to reference thin sections of ethnoarchaeological and experimental floors, including those from the floors of Iron Age houses reconstructed by the late Dr Peter Reynolds at Butser Ancient Farm, Hampshire, UK. This paper reports upon experimental studies mainly carried out at Butser during 1990–95 and their suggested application to microstratigraphic and microfacies analysis of occupation sites. Sampling at Butser in 1990 was very much a ‘rescue’ operation. Nevertheless, fundamental differences in soil micromorphology were found between domestic house beaten floor deposits and those formed in a stable. These differences were also reflected in the parallel study of pollen from these deposits, with microchemical analyses helping to explain anomalous pollen preservation in the ‘unsuitable’ soil at Butser. The limited bulk chemical study produced less clear results, but when compared with findings from some Roman and early medieval sites in London, showed that the chemical trends hinted at Butser are probably replicated in some ancient floors. A preliminary interdisciplinary study of samples from the London Guildhall clearly suggested the potential of such microstratigraphical and microfacies analyses, and intimates at the potential of attaining rigorous consensus interpretations. Stable floor deposits are both organic and phosphate-rich, with homogeneous soil micromorphological and pollen characteristics. Beaten floors are much more mineralogenic and heterogeneous in terms of their microfabric and pollen that also indicate trampling-in of a wide variety of allocthonous materials, the last commonly responsible for raised magnetic susceptibility at archaeological sites. These findings prove the worth of the ethnoarchaeological simulations carried out at the Moel-y-gar stable and domestic Pimperne House. This microfacies approach is viewed as complementary to site reconstruction through macrofossil analysis of beetles and plant remains, especially where sites are being investigated at the microstratigraphic level. The phosphate content of stabling deposits may allow preservation of pollen in what may otherwise be deemed unsuitable material, as demonstrated at Butser.