Purpose: To determine the criterion validity of the Physical Activity Scale for Individuals with Physical Disabilities (PASIPD) to assess the level of everyday physical activity (PA) in persons with a physical disability. Methods: In a total of 124 ambulatory and non-ambulatory subjects with cerebral palsy, meningomyelocele or spinal cord injury, we assessed level of everyday PA using the PASIPD, a recall questionnaire, and using an accelerometry-based activity monitor (AM), which served as reference method. Results: Significant (p ≤ 0.05) spearman correlation coefficients between the outcome parameters of the PASIPD and AM ranged between 0.22 and 0.37. The PASIPD significantly (p<0.0001) overestimated the duration of PA (3.9 (2.9) versus 1.48 (0.90) hours per day by AM). There was a significant correlation (spearman's rho -0.74, p<0.0001) between the level of everyday PA and the discrepancy between the methods, indicating higher overestimation with increasing level of everyday PA. Conclusion: The PASIPD is poorly correlated with objective PA measurements using the AM in people with a physical disability. Although similar low correlations between objective and subjective PA measurements have been found in the general population, users of the PASIPD should be cautious for overestimating physical activity levels.