Abstract: Adult males, females or fourth instar larvae of the ladybird Harmonia axyridis were starved for 24 hours and then released in a 9 cm Petri dish with either 20 females of one of the aphids: Acyrthosiphon pisum, Aphis fabae cirsiiacanthoidis or Aphis sambuci, or 10 + 10 females of two aphid species combined. After 6 hours, we scored the number of aphids eaten completely, eaten partially or that had their bodily contents sucked out. Males consumed generally the same number of aphids as females, but sucked out aphid bodily contents without eating the prey cuticle more often than females. In no-choice experiments, adults consumed more A. pisum than A. f. cirsiiacanthoidis and the smallest number of A. sambuci, which is slightly toxic. In choice experiments, neither males nor females preferred either of the two aphids in any of the treatments, except for males eating more of the toxic A. sambuci than A. f. cirsiiacanthoidis. Larvae ate more of the suitable food A. f. cirsiiacanthoidis than the toxic A. sambuci. When we presented aphids on either a black or green background for the choice experiment, females still did not prefer any aphid species, except for black A. f. cirsiiacanthoidis being consumed more often than green A. pisum on a black background. Because ladybirds consumed much more A. pisum in no-choice experiments on both backgrounds, we should compare the ratio observed in choice experiments with the reversely biased ratio in no-choice experiments, not with a 50:50 probability. Thus, the difference in the choice experiment is very strong and its polarity surprising.