Primary and secondary cultures of rhesus monkey kidney cells supported multiple-cycle replication of Sendai virus, but later passages lost this ability, and this was reflected in decreased plaque formation. Multiple-cycle replication also did not occur in LLC-MK2 cells, a continuous line of RMK cells. Failure of replication in serially passed cells was correlated with a decrease in proteolytic cleavage of a viral surface glycoprotein (Fo), and the ability of cells to support multiple-cycle replication and plaque formation could be restored by the addition of trypsin (0.3 microgram/ml) to the overlay medium. The use of wild-type virus, which requires trypsin, and protease activation mutants that require chymotrypsin or elastase for activation has provided evidence that the activating protease supplied by primary or secondary cells has trypsin-like activity. Inactive virus, with uncleaved Fo glycoprotein, absorbed to primary or secondary cells but did not infect them, even though such cells possess the enzyme that is capable of cleaving the Fo glycoprotein of virus synthesized in these cells. The inability of these cells to activate adsorbed virus indicates that the activating protease that they possess is inacessible to adsorbed virus, although it can act on the Fo glycoprotein during virus maturation in these cells. These data provide a biochemical explanation for the failure of later passages of a cell strain or a continuous cell line to support the replication of a paramyxovirus.