A spatial-averaged model of boiling flow in an electrically heated vertical pipe is used to investigate and explain experimental data of various labroatories. The agreement with data is good over a variety of conditions ranging in pressure from atmospheric to 1000 psia, and in heated length from 2 to 16 ft. Two slip-ratio correlations are compared in testing the model against the stability data; the correlation of Bankoff being less successful at low subcooling than the modified Bankoff correlation due to Jones. A value of Bankoff's K recommended by Kholodovski is also compared for Spigt's experiment. The crucial boiling length, where the system is least stable, is used to demonstrate the dependence of stability on heating rate, flow rate, and degree of subcooling. In particular, it is shown that with the Bankoff-Jones slip ratio, an increase in the ratio of heating rate to flow rate invariably leads to less stable conditions. On the other hand, an increase in subcooling leads to less stable conditions only when the degree of subcooling is less than that at the crucial boiling length.