The International Classification of Functioning, Disability and Health (ICF) has potential to be used as a clinical assessment instrument directly. However, difficulty in operationalizing the ICF qualifiers has led to unsatisfactory inter-rater reliability of ICF instruments in previous studies. The ICF set for spinal cord injury (SCI) nursing contains a group of categories that can reflect the functioning of SCI patients from the perspective of nurses. The aim of this study was to explore the inter-rater reliability of the ICF set for SCI nursing. Detailed measuring guidelines were prepared for each category of the ICF set. Two trained nurses then used the ICF set to independently rate 40 SCI patients within the first 3 days after their admission. The results showed that the percentage of the observed agreement between the nurses ranged from 42.5 to 100% (median 75%, interquartile range 62.5–87.5%). The weighted κ ranged from −0.03 to 1.00 (median 0.68, interquartile range 0.45–0.84). A total of 50 categories (79.4%) showed weighted κ greater than 0.4 and 39 categories (61.9%) had weighted κ greater than 0.6. The medians of the weighted κ for the body functions, body structures, activity and participation and environmental factors components were 0.6, 0.64, 0.84, and 0.11, respectively. This study indicated that the inter-rater reliability of the ICF set for SCI nursing was acceptable. Establishment of detailed measuring guidelines could help reduce the differences between raters. Simpler and clearer measuring guidelines are recommended and the definitions of some categories need to be clarified in future studies.