The formation of methane hydrate beds at the gas-water interface in a high-pressure visual autoclave apparatus, under both continuous cooling/flow and shut-in/restart operating procedures, was studied. Bed formation was identified by an increase in the measured resistance-to-flow of the hydrate slurry, and supported by visual observations. During continuous cooling/flow experiments, the hydrate volume fraction required to form a moving bed increased from 15 vol % to 40 vol % over a range of initial Reynolds numbers for the stirred cell of 280 to 4500. For shut-in/restart trials, the bed formation point increased from 6.6 vol % to 33 vol % hydrate over an equivalent, stirred cell Reynolds number range of 240 to 3900. No significant differences in the dependence of the bed formation point on shear rate were observed between the constant cooling/flow and shut-in/restart experiments, suggesting both systems evolved along the same pathway to hydrate plug formation. Some differences between the two types of experiments were observed in the dependence of the initial hydrate formation rates on Reynolds number, with shut-in/restart experiments having formation rates up to an order of magnitude larger. More significantly, in both types of experiments the formation rate increased logarithmically with Reynolds number. The dependences of bed formation and growth rate on shear are crucial results for assessing the risk of forming a hydrate plug in mature production systems in which water is the dominant phase. High shear increases hydrate growth rate, but delays the onset of hydrate bed formation, which is the precursor to plugging. By trading-off these competing effects, it may be possible to develop an optimum restart strategy to minimize the risk of hydrate plug formation.