Odor memory is commonly believed to be very strong and long-lasting. The present study examined factors that impact odor recognition memory over short delay intervals (immediately or 30 s after target presentation) with emphasis on memory task (forced-choice vs "monadic"/single stimulus yes/no), odor category, and target/foil relationship. We explored trial-by-trial confidence as well as the effect of target familiarity, pleasantness, and intensity ratings, and odor nameability on memory for odors. Overall odor recognition memory in terms of proportion correct and sensitivity measures did not decline significantly during the 30-s delay interval in either task. However, hit rates were lower at 30 s and correct rejection rates for common odors remained consistently high. Recognition memory was better on trials in which the odor pairs were highly dissimilar, as well as on trials in which the target was an uncommon odor, particularly if it could be named. Familiarity, pleasantness, and intensity had no systematic effect on recognition memory. Whereas the results provide evidence of a fading memory trace, indicated by the decreased hit rates after a 30-s delay, the constant rates of correct rejections and high confidence ratings on those trials, even after delay, suggests that novelty detection (i.e., recognition that an odor is not one that has been encountered previously in that context) may play an important role in the memory for odors over short delays. Whether there is a separate short-term odor memory store is also addressed.