In this article Power Spectral Density Analysis of the velocity profile of handwriting tasks is applied to estimate movement noise in poor and proficient writers. It is hypothesized that poor writers are less effective in the inhibition of neuromotor noise. Evidence for this theory is found in an analysis of electronically sampled handwriting movements of 48 children from grade 2, 3 and 4 (mean ages 8, 9 and 10, resp.). Twenty-four were poor and 24 were good writers, as judged by their teachers. Subjects were matched on age, sex, handedness and educational level. They performed handwriting tasks consisting of simple garlands and arcades, or meaningless words built from the letters 'e', 'n' and 'm'. In the tasks stroke accuracy, size and rotational direction were systematically varied. The results reveal that movement times were not different between the two groups, but poor writers and good writers used different movement strategies. Movements of the less proficient children were larger and were produced with higher movement velocities. Power Spectral Density Analysis disclosed that handwriting movements of poor writers were substantially more noisy than those of proficient writers, with a consistent peaking of the noise energy in that region of the spectrum which is representative for neuromotor tremor. Also, poor writers were less successful in adapting the level of noise to increased accuracy demands of the tasks. The results support the view that deviant biomechanical strategies might be responsible for deficient motor performance.