Abstract The biological activity of the Ireland strain of Choristoneura fumiferana (Clem) nucleopolyhedrovirus ( Cf MNPV) propagated in different hosts was determined to provide the basis upon which genetically modified Cf MNPV, or other naturally occurring isolates, should be compared. Occlusion bodies (OB) derived from CF‐203 cells were significantly larger and more pathogenic than those propagated in vivo when tested against the fifth larval instar of C fumiferana (Clem) and C occidentalis Freeman. The dose‐responses (LD 50 and LD 95 , expressed as occlusion bodies per larva) of C fumiferana larvae to in vitro ‐propagated OBs were 274 and 5785, respectively. The values of LD 50 and LD 95 to C occidentalis larvae were 19 and 118, respectively. There were no significant differences in pathogenicity or size when OBs propagated in C fumiferana larvae were tested against either insect species, nor were there significant differences for OBs propagated in C occidentalis larvae. The LD 50 and LD 95 of in vivo ‐produced OBs to C fumiferana were 925 and 61 988, respectively. The LD 50 and LD 95 to C occidentalis were 50 and 453, respectively. OBs propagated in vitro had a mean volume of 13.13 µm 3 , whereas those propagated in vivo ranged from 0.84 to 1.41 µm 3 . The median survival time‐responses (ST 50 ) of fifth‐instar C fumiferana or C occidentalis larvae to OBs propagated in vivo were not significantly different from those propagated in vitro at the dosage levels tested. Values of ST 50 of C fumiferana larvae to in vitro ‐ and in vivo ‐produced OBs at dosages causing less than 50% mortality ranged from 9.6 to 9.8 days post‐inoculation (dpi), whereas a LD 95 dose resulted in ST 50 values ranging from 7.3 to 7.7 days. ST 50 values of C occidentalis larvae at dosages causing less than 50% mortality ranged from 9.8 to 10.2 dpi, whereas a LD 95 dose resulted in ST 50 values ranging from 9.5 to 9.8 dpi. The median feeding cessation time‐response (FT 50 ) of fifth‐instar C fumiferana larvae to OBs propagated in vitro (5.7 days) was not significantly different from the FT 50 of those propagated in vivo in either insect species (5.3 and 5.7 days) at the dosage level tested (LD 95 ). No significant differences in FT 50 values were observed between OBs propagated in either larval host. The FT 50 of C occidentalis larvae to OBs propagated in vitro (7.7 days) was not significantly different from that to those propagated in vivo in C occidentalis larvae (7.6 days), but somewhat different (7.2 days) from that to those propagated in C fumiferana larvae. Results indicate that Cf MNPV can be propagated in vivo in either C fumiferana or C occidentalis larvae (or sequentially through both) without alteration in infectivity, although the use of the CF‐203 cell line yields the most biologically active OBs. Copyright © 2004 Society of Chemical Industry