The use of on-line relative permittivity (epsilon＇) measurements for monitoring cultures of Sf-9 cells was evaluated in a batch culture and a batch infected with a baculovirus expressing beta-galactosidase. It was found that viable cell density a nd volume essentially accounted for all the variation in epsilon＇ in both non-infected and synchronously infected cultures, indicating that the epsilon＇ of a cell suspension was sensitive only to changes in the viable cell population. Additionally the parameter provided clearly defined signposts of the progress of the infection.