The presence of tissue factor (TF) enhances the activity of factor VIIa (FVIIa) toward FX by a factor of approximately 10 5 if full length [1,2] and by a factor of approximately 76 in the soluble (TFs), truncated form [3]. The activation of FVII is also greatly enhanced by the presence of TF [4]. As the initial X-ray crystal structure of the complex of TFs/FVIIa/inhibitor revealed that TF does not bind at or near the active site of FVIIa [5], there has been a substantial interest in the question: how is the allosteric effect of TF (or TFs) on FVIIa manifested? We have previously built equilibrated solution models of FVII zymogen [6], FVIIa/TFs [7] and TFs/FVIIa/FXa [8] so as to extend the X-ray data into the realm of short-time dynamics (less than a few nanoseconds) to investigate local changes. Likewise, others have provided mutational studies that provide clues for the underlying nature of the allosteric effect [3,9]. Recently, it has been proposed [10] that FVII/FVIIa exists in two forms, one of which is competent to be enhanced by TF, and the other not. The existence of these two forms was suggested [10,11] as a result of a structural study in which it was found that FVII zymogen (serine protease + egf2 domains) has a local conformational change -a novel registration of b-sheets B2-A2 -not seen in the explicit FVIIa structures. The competent form was taken to be that seen in the original FVIIa/TFs structure; the incompetent form was taken to be that seen in the FVII zymogen structure [10,11]. Additionally, the zymogen structure has a bound non-active site inhibitor; this inhibitor has also been shown to block the binding of TF to FVIIa [12]. The zymogen structure [10] inspired others to test the hypothesis by designing mutations that blocked the shift between the two proposed forms in FVIIa [13]; the results of these experiments did not substantiate the FVIIa two-state hypothesis as the mutations appeared to have little effect on FX cleavage rates.The publication of the structural data on the FVII zymogen [10] that led to the two-state hypothesis for FVIIa compares the FVIIa/TF structure and the FVII zymogen structure. Only one set of crystal contacts is defined [10], these contacts being away either from the inhibitor (A183) binding site or that of local conformation change. When we did (by several independent codes) a tally of all of the crystal contacts (pdb ¼ 1JBU), we were surprised to find that a second molecule in the unit cell (four molecules per unit cell) contacts the first molecule, the points of contact being through several regions of the zymogen of the first FVII molecule with the inhibitor bound to the second zymogen FVII molecule. Several contacts are strong hydrogen bonds. What is vitally important for this discussion is that the point of contact on the first molecule occurs at precisely the location of the local conformational change observed in the structural work for FVII. The interaction of the peptide bound to the second zymogen molecule with the first zymogen is clearly seen in Fig. 1a...