Test-retest reliability of a pain drawing instrument was investigated. Pain drawings of chronic pain patients (n = 51) were scored for percentage of total body surface in pain and location of pain. A test-retest reliability coefficient of r = 0.85 was calculated for a time interval that averaged 71 days. In addition, a percentage of agreement based on distribution of pain over time was calculated at 88.2%. The effect on reliability of age, gender and time-interval differences was investigated. The utility of the pain drawing instrument as a measure of extent of pain and location of pain over time is discussed.