Trajectories of diatomic ligands in heme proteins have provided a wealth of information on how such molecules can enter and exit the protein, which must fluctuate in order to permit their passage. The time-resolved X-ray crystallography technique has provided the most detailed images of the transport of the carbon monoxide ligand (CO) to various cavities within proteins (1–5). However, the function of protein cavities is still an open question. Indeed, it is difficult to assign a general function based on the analysis of a single protein, since the functional role of cavities may vary from one protein to another. Thus far, two heme proteins, Sperm Whale myoglobin (SWMb) (2–3) and Scapharca inaequivalvis dimeric hemoglobin (HbI) (4–5), have been the most extensively studied by time-resolved crystallography. Although one is a myoglobin and the other a hemoglobin they both have the reversible binding of O2 as their primary function. Despite the rich set of conformational changes associated with the allosteric cooperative transition in photolyzed HbI*CO, the nature of the CO trajectory in that globin has certain similarities with SWMb*CO (The asterisk indicates a photolyzed carbon monoxide molecule, following photon absorption by the heme that leads to rupture of the Fe-CO bond). Both have a distal docking site B, where the initially photolyzed CO molecule resides. The B-site in both proteins is a distinct ligand docking site in the distal pocket, not observed as the Xe binding site in either proteins. (2, 4) Based on experiments designed to block the Xe-binding cavities in SWMb*CO and HbI*CO, and on time-resolved crystallographic studies, it has been deduced that the cavities in these proteins do not constitute an exit route for diatomic ligands (6–7). Instead, exit occurs near the distal histidine in both SWMb*CO and HbI*CO. Moreover, previous temperature derivative spectroscopy (TDS) studies show the presence of a nearby docking site and one or more further secondary sites in SWMb (8–10) and HbI (11) where photolyzed CO molecules can reside. Why then does CO migrate to a number of Xe-binding cavities in these proteins after it leaves the B-site? This fundamental question has been studied extensively in SWMb using many spectroscopic techniques including TDS and methods such as kinetic hole burning (12), which makes the connection between the conformation-dependent energy of recombination and the spectroscopic energy of a heme charge transfer band. In this study we propose that in order to understand the possible function of the cavities in various heme proteins, it is necessary to compare a protein with a significant difference in function. We conclude that ligand dynamics studied by time-resolved X-ray and the protein cavity in DHP identified by a Xe binding site (13) are consistent with an open architecture in the distal pocket of the bifunctional hemoglobin, DHP, which explains how it can carry out multiple functions. These features distinguish DHP from the more specialized oxygen transport proteins SWMb and HbI. Dehaloperoxidase-hemoglobin (DHP), first isolated from the terebellid polychaete Amphitrite ornata, is a bifunctional protein, which displays significant peroxidase activity under physiological conditions, while also having a globin fold and an associated oxygen transport function (14). As a hemoglobin, DHP can reversibly bind with O2 and as a peroxidase, DHP can oxidize 2,4,6-trihalophenol into the corresponding 2,6-dihaloquinone in the presence of H2O2. However, normal peroxidase function inactivates the ability of DHP to bind oxygen leading to a paradox because of the failure of the DHP hemoglobin function. (15) Recently, this functional paradox of DHP has been resolved by establishing that oxyferrous DHP can serve as the resting state of the peroxidase reaction cycle. (16–17) For a typical peroxidase, e.g. Horseradish Peroxidase and Cytochrome c Peroxidase, the ferric state of the heme Fe is the resting state. Thus, DHP appears to have a new type of peroxidase cycle, a ferrous peroxidase cycle. (15) Although it has high peroxidase activity, the oxyferrous peroxidase cycle and unique internal binding sites possessed by DHP distinguish it from other heme peroxidases. (18–19) The replacement of O2 bound to the heme Fe by H2O2 required by the ferrous peroxidase mechanism suggests DHP must simultaneously accommodate both O2 and H2O2 in the distal pocket in order to function. This requirement is quite different from other well-studied hemoglobins and myoglobins. For example, SWMb, which is by far the best studied protein, cannot simultaneously accommodate both CO and H2O in the distal pocket (20–22). The reasons for this may become clearer in comparison with the bifunctional protein, DHP, which has a different requirement for the function of the distal pocket, namely entry of multiple molecules both for activation and for inhibition of enzymatic function. In order to address the dynamics of diatomic ligands in DHP, we have conducted a time-resolved X-ray crystallography study of the trajectory of photolyzed carbon monoxide in DHP*CO. The time-dependent structures reveal a qualitatively different CO trajectory than those observed previously in SWMb*CO and HbI*CO.