Laboratory flume experiments were carried out, to measure the effect of biota on erodibility of mudflat sediments. The experiments sought to reproduce the environment of the lower mudflat at Hythe, Southampton Water, Southern England; this is characterised by fine grain-size and a surface layer of very fluid mud. Natural sediments were used to produce settled beds in the Lab Carousel, an annular flume of 2 m diameter. The following bed conditions were investigated diatom biofilms; the addition of cockles (Cerastoderma edule); and abiotic sediment, obtained by the addition of sodium hypochlorite. The erosion threshold (?crit, calculated with the TKE method) was in the range 0.02–0.20 Pa. Bioconsolidation increased ?crit considerably: compared to the abiotic sediment experiment, ?crit was 5–10 times higher depending on the biofilm development. The relationship between ?crit and water content of sediment (the best proxy for sediment compaction) was as good, or better than between ?crit and chlorophyll a (proxy for biofilm development). When cockles were introduced, ?crit was significantly lower (reduction by 50–75% compared with the diatom biofilm experiments), reflecting the surface disturbance by the bivalves. The biofilm erosion was characterised by a patchy pattern: the bed surface stayed mainly uneroded and erosion was visible only on a few elongated patches commencing at some weakness points of the biofilm, then progressing downstream. The results illustrate the importance of the surface heterogeneity: the irregularities of a natural bed (weak points of the biofilm, bioturbations, microrelief, larger roughness elements like shells or algae, etc.) have a determinant effect on the erodibility of biofilms. Such characteristics may have more influence than biofilm strength, because the erosion starts from the weaker areas.