We examined participants' abilities to manually estimate one of two perpendicular line segment lengths using curved point-to-point movements. Configurations involved symmetrical, unsymmetrical, and no bisection in upright and rotated orientation alterations to vertical-horizontal (V-H) illusions, where people often perceive longer vertical than horizontal segments for equal segment lengths. Participants used two orthogonally directed movements for length estimations: positively proportional (POS) - where greater fingertip displacement involved longer length estimation between configuration intersection start position and fingertip end, and negatively proportional (NEG) - where greater fingertip displacement from the screen edge start position toward configuration intersection involved a shorter length estimation between configuration intersection and fingertip end. Length estimations followed most standard perceptual aspects of the V-H illusion for POS estimations, yet differed between upright and rotated orientations for the symmetrical configuration. NEG estimations revealed no illusory influences. Use of allocentric programming likely accompanied POS estimations to explain V-H illusory influences on perceptuomotor control.