A need to consider possible gender differences in pain research has been recognized by researchers during the last decades. As part of a psychometric evaluation of the Swedish version of the Multidimensional Pain Inventory (MPI-S), we performed gender-differentiated analyses of the internal consistency, validity and sensitivity to change of the MPI-S in a sample of 235 individuals (129 females, 106 males) suffering from long-term non-specific pain from the lower back and/or neck region. The construct validation and sensitivity analyses were performed by using validated self-report measures and direct observational assessment techniques as external constructs. For sections 1 and 2 of the MPI-S, the results support the internal consistency (alpha coefficients ranged from 0.74 to 0.85 for females and 0.62 to 0.89 for males) and construct validity across gender. The General Activity (GA) scale of section 3 of the MPI-S displayed acceptable internal consistency across gender (alpha = 0.79 for females, 0.80 for males) but not a satisfactory construct validity. Furthermore, the results yielded some support for the sensitivity to change of the Pain Severity (PS), Interference (1), Life Control (LC) and Affective Distress (AD) scales (from section 1) across gender. Unfortunately, the GA scale did not display a satisfactory sensitivity either for females or males. Altogether, the results showed a similar pattern across gender, although some divergences were detected, such as the substantially weaker negative correlation between perceived supportive behaviour from significant others and punishing responses for males compared to females. In conclusion, we recommend the use of sections 1 and 2 of the MPI-S as a psychometrically evaluated and comprehensive instrument in the assessment of individuals suffering from chronic non-specific low back pain or neck pain.