An "outlier" in a frequency distribution refers to a data point that significantly differs from the other values in the dataset. It is an unusually high or low value that can distort statistical analyses and may indicate variability, errors, or unique occurrences.