Studies of blocking between elements from the same stimulus modality elucidate how animals analyze such mixtures. Proboscis-extension conditioning of honey bees was used to examine the role of unconditioned stimulus (US) intensity in blocking in odorant mixtures. Several predictions from theoretical accounts of blocking were tested. One-trial blocking was found when subjects were pretrained with a high but not a low US concentration. Increases in US intensity produced unblocking, but decreases did not. Furthermore, an equivalent level of unblocking was observed when the increase occurred on the 1st or 2nd mixture-phase trials. These results are consistent with studies of vertebrates in which US intensity is altered to produce unblocking. Extension of these studies to include variation in other US parameters and to intramodal blocking in vertebrates will be necessary to determine whether this correspondence to vertebrate blocking is robust.