ABSTRACT: BACKGROUND: In self-renewing, pluripotent cells, bivalent chromatin modification is thought to silence (H3K27me3) lineage control genes while "poising" (H3K4me3) them for subsequent activation during differentiation, implying an important role for epigenetic modification in directing cell fate decisions. However, rather than representing an equivalently balanced epigenetic mark, the patterns and levels of histone modifications at bivalent genes can vary widely and the criteria for identifying this chromatin signature are poorly defined. RESULTS: Here we initially show how chromatin status alters during lineage commitment and differentiation at a single well-characterised bivalent locus. In addition we have determined how chromatin modifications at this locus change with gene expression in both ensemble and single cell analyses. Next we show, on a global scale, how mRNA expression may be reflected in the ratio of H3K4me3/H3K27me3. CONCLUSIONS: While truly "poised" bivalently modified genes may exist, the original hypothesis that all bivalent genes are epigenetically pre-marked for subsequent expression might be over-simplistic. In fact from the data presented here, it is equally possible that many genes, which appear to be bivalent in pluripotent and multipotent cells, may simply be stochastically expressed at low levels in the process of multi-lineage priming. Although both situations could be considered to be forms of "poising", the underlying mechanisms and the associated implications are clearly different.