Two experiments are reported that examine the effects of morphological overlap between prime and target in the masked priming paradigm. Experiment 1 demonstrated that the prior presentation of a higher frequency morphologically related word facilitates target processing. This facilitation obtained only to prefixed and not to suffixed targets and was independent of whether the prime was a stem or another prefixed form. In Experiment 2 morphological priming was estimated using both an unrelated and an orthographically related baseline. Approximately equivalent facilitation was observed for prefixed and suffixed primetarget pairs when measured against the orthographic control. The results suggest that the effects observed in masked morphological priming reflect the combination of both facilitatory and inhibitory mechanisms.