Previous studies (Mertens et al., Virology 157, 375-386, 1987) have shown that removal of the outer capsid layer from bluetongue virus (BTV) significantly reduces (approximately x 10(-4)) the infectivity of the resultant core particle for mammalian cells (BHK 21 cells). In contrast, the studies reported here, using a cell line (KC cells) derived from a species of Culicoides that can act as a vector for BTV (Culicoides variipennis), demonstrated a much higher infectivity of core particles than that in mammalian cells (approximately x 10(3)). This increase resulted in a specific infectivity for cores that was only 20-fold less than that of purified disaggregated virus particles (stored in the presence of 0.1% sodium-N-lauroylsarcosine (NLS)). Removal of this detergent caused intact virus particle aggregation and (as previously reported) resulted in an approximately 1 log10 drop in the specific infectivity of those virus particles which remained in suspension. In consequence the specific infectivity of core particles for the KC cells was directly comparable to that of the intact but aggregated virus. These data are compared with the results from oral infectivity studies using two vector species (C. variipennis and Culicoides nubeculosus), which showed similar infection rates at comparable concentrations of purified cores, or of the intact but aggregated virus particles (NLS was toxic to adult flies). The role of the outer core proteins (VP7) in cell attachment and penetration, as an alternative route of initiation of infection, is discussed. Previous studies (Mertens et al., Virology 157, 375-386, 1987) also showed that the outer capsid layer of BTV can be modified by proteases (including trypsin or chymotrypsin), thereby generating infectious subviral particles (ISVP). The specific infectivity of ISVP for mammalian cells (BHK21 cells) was shown to be similar to that of disaggregated virus particles. In contrast, we report a significantly higher specific infectivity of ISVP but not of the intact virus (approximately x 100) for two insect cell lines (KC cells and C6/36 mosquito cells (derived from Aedes albopictus)). In oral infection studies with adults of the two vector species, ISVP produced the same infection rate at approximately 100-fold lower concentrations than either core particles or the intact but aggregated virus particles. The importance of mammalian host serum proteases, or insect gut proteases, in modification of the intact virus particle to form ISVP and their role in initiation of infection and the vector status of the insect is discussed.