OBJECTIVES: The Self-Assessment of Treatment Version II (SAT-II) instrument has three versions: baseline; follow-up (FU); and last-visit versions. Baseline version measures pain level (one item) and impact of pain (six items). Follow-up and last-visit versions measure improvement in the impact or level of pain  due to treatment. Last-visit also includes two items about treatment preferences. This study explored scoring of the SAT-II (baseline and last-visit). Scores were also assessed for their responsiveness to change. METHODS: Data from 369 patients (42% female; mean age=63) with painful diabetic peripheral neuropathy from a randomised, double-blind placebo-controlled trial (STEP) were used to test baseline and last-visit versions. Item-to-item spearman correlations and spearman correlations between summary scores and Brief Pain Inventory-Diabetic Neuropathy pain severity score (PSS) were tested. Factor analysis was conducted on the six impact items to develop a scoring algorithm (baseline or last-visit summary score). Analysis of variance measured responsiveness to treatment responder status as defined by PSS (both ≥30% & ≥50% reduction from baseline) and Patient Global Impression of Change (PGIC; much/very much improved). RESULTS: A single-factor structure for the six impact items was demonstrated for baseline and last-visit. Item-to-item correlations (baseline: correlation (r)=0.33 to 0.75; last-visit: r=0.52 to 0.89) and correlations between a sum of the impact items and PSS were significant for baseline (r= 0.39, p=0.001) and last-visit (r=-0.46, p=0.0001). A mean summary score of the impact items discriminated (all p<0.0001) between non-responders and responders, respectively for each definition as follows: ≥30%, 0.6 (SD 0.9) vs 1.9(1.3); ≥50%, 0.8 (1.0) vs. 2.3 (1.3); and PGIC, 0.6 (0.7) vs. 2.3 (1.2). CONCLUSIONS: The six impact items of the SAT-II can be scored as a summary measure on both the baseline and last-visit versions, whilst other items should be scored individually. The last-visit version discriminates treatment response.