This study compared the consistency of the Short Form (SF) and the Long Form (LF) of the Bruininks-Oseretsky Test of Motor Proficiency (BOTMP) in identifying preschool children with motor impairment (MI). One hundred and forty-four Greek preschool children participated (74 males, 70 females; mean age 5 y 2 mo [SD 5 mo], range 4 y 6 mo-5 y 6 mo). Although total SF and LF scores were highly correlated (r=0.85), paired t-tests indicated significant differences (t=-27.466, p=0.001). SF total scores (mean 58.72 [SD 7.28]) were higher than LF total scores (mean 47.38 [SD 9.43]). SF had low sensitivity (13.6%) and negative predictive value (72.5%) for identifying MI. The BOTMP-SF does not appear to be a valid test for the identification of MI in 5-year-old children.