The level of agreement between clinicians' ratings of gait patterns in children with spastic hemiplegia was investigated using the Winters, Gage, and Hicks (WGH) classification scale. Sixteen clinicians (nine physiotherapists, seven orthopaedic surgeons) from six gait laboratories in five different countries rated gait patterns in 34 children with spastic hemiplegia (22 males, 12 females; mean age 10y [SD 3y], range 6-18y). Archived three-dimensional gait reports (kinematic and video data) and videos (video clips only) were rated for each child. Agreement between clinicians was substantial for reports (weighted kappa [wkappa] 0.77, range 0.62-0.89). Agreement was lower for videos (wkappa 0.63, range 0.39-0.97) and for clinicians' agreement between reports and corresponding videos (wkappa 0.62, range 0.47-0.76). Exact agreement was unacceptable for some gait patterns using reports (mean 65%, range 32-74%) and videos (mean 53%, range 35-94%). Not all gait patterns could be rated using the original WGH categories. It was concluded that: (1) agreement is acceptable using the WGH tool; (2) kinematic data from 3D instrumented gait analysis and video should be used together when using the WGH scale; and (3) further refinement of this classification is required.