Research among persons with cancer pain suggests that the association between pain intensity and pain interference is non-linear. That is, pain begins to have a serious impact on functioning when it reaches a certain threshold level (about 5 on 0--10 scales). Often, a second pain threshold can be identified which, once reached, shows an even greater proportional negative impact on functioning. This finding supports the potential clinical utility of classifying pain as mild, moderate, and severe based on the impact of pain on quality of life, and research among persons with cancer pain supports specific cutoffs (mild: 1--4, moderate: 5--6, severe: 7--10, see Pain 61 (1995) 277) for this classification. The current study sought to replicate the non-linear association between pain and pain interference in a non-cancer pain sample, determine whether the cutoffs that have been identified as optimal for cancer patients are also optimal for persons with pain associated with amputation, and determine whether the optimal cutoffs replicate across pain types (in this case, phantom limb, back, and general pain) within a single sample. Two-hundred and five persons with acquired amputation and phantom limb pain, back pain, or both, rated their average pain intensity and degree of pain interference for each type of pain. The results support a non-linear association between pain intensity and pain interference. However, the optimal cutoffs for classifying mild, moderate and severe pain in the present sample replicated the findings for persons with cancer pain only for back pain -- different optimal cutoffs were found for phantom limb and general pain. Moreover, the degree of pain interference appeared to vary as a function of pain type. The same level of back pain interfered more significantly with daily function than phantom limb pain did after pain levels reached five or more (on a 0--10 scale). These findings have implications for understanding the meaning of pain intensity levels, as well as for the assessment of pain intensity in persons with amputation-related pain.