We have developed an experimental assay to monitor the rate of dissociation of the T4 DNA polymerase holoenzyme (polymerase plus gene 44/62 and 45 proteins) once it has been stalled by nucleotide omission. Using this assay, we determined that the dissociation of the DNA polymerase holoenzyme follows a first order decay with a half-life of 2.5 min. The long half-life resembles that expected for the holoenzyme processively synthesizing DNA on the leading strand of the replication fork. The holoenzyme dissociation rate is independent of polymerase accessory protein concentration and of ATP hydrolysis. The dissociation rate is increased if the gene 32 protein is omitted or if the primer-template region is shortened from 46 to 28 base pairs. But the rate of holoenzyme dissociation is most strikingly increased when the circular DNA template is converted to a linear form. By analogy with other well studied systems, these results support a model in which ATP hydrolysis by the 44/62 proteins serves to load a ring-like 45 protein onto the DNA. Once loaded, the 45 protein, possibly along with the 44/62 complex, acts as a sliding clamp that tethers the DNA polymerase to the template.