Convergent validity represents the extent to which a test demonstrates ability to measure a theoretical construct (Beaton et al. 2000). The construct validity of the Turkish version SF-MPQ was assessed by principal component analysis. For this purpose, it was correlated with NRS that has been widely used clinically for the assessment of pain (Paice & Cohen 1997, Childs et al. 2005) obtained on the same day. Correlations above 0Æ40 have been considered satisfactory for convergent validity (Kaasa et al. 1995). We found statistically significant and positive between the Turkish version of the scale and NRS values (r = 0Æ61 for test and r = 0Æ68 for retest), indicating good validity. Our results are supported by Yakut et al. (2007) who used NRS as a construct validity criterion in their study, and reported correlation coefficients similar to those described in this study (r = 0Æ63 for test and r = 0Æ70 for retest). This result could be due to the fact that the patients come from the same cultural background.