The authors have measured the amounts of interferon formed by chick cells 'aged' in vitro in response to different amounts of infectious wild-type Sindbis virus. Their results suggest that one plaque-forming unit is enough to induce maximum interferon formation. With higher m.o.i. the yield of interferon is less. To inactivate the interferon-inducing activity of Sinbis virus, four times more u.v.-radiation was needed than to inactivate the infectivity of the virus. This suggests that only 25% of the virus genome need be intact in order to induce interferon. Temperature-sensitive Sindbis virus mutants from the three RNA+ complementation groups, C, D and E, gave rise to interferon in chick cells incubated at a non-permissive temperature. Similarly, mutants from two of the RNA- groups, B and F, gave rise to interferon, but not mutants from groups G and A. They conclude that no pre-formed inducer of interferon is present in Sindbis virus. It appears, however, that genes G and A represent a special one-quarter of the genome which must be functional in order to synthesize an interferon-inducing moiety. They suggest that this moeiety is a double-stranded RNA molecule formed after synthesis of a segment of RNA complementary to the genome.