Handwriting (HW), unlike reaching or walking, is a high-level motor activity, engaging large parts of cortical and sub-cortical regions that include supplementary motor area(SMA), premotor area(PM), primary motor area(M1), basal ganglia(BG), cerebellum, spinal cord etc. Since each of these regions contributes to HW output in its own unique fashion, pathology of any of these regions is manifest as characteristic features in HW. For example, in Parkinson's disease, a disorder of BG, HW is marked by diminutive letter size or micrographia. Recognition of rich diagnostic value of HW had prompted a systematic study of HW and the extensive neuromotor organization that generates it. Computational modeling offers an integrative framework in which results of such studies, which come from several domains, like behavioral, imaging, etc are brought together and given a concrete shape. An integrative computational model of human motor system and BG is proposed. Dopamine deficient conditions as in PD patients are simulated in the model to reproduce PD-like handwriting features like micrographia, fluctuating velocities, jagged contour etc. The model primarily consists of a neuromotor model which is capable of learning and generating learnt strokes, and a timing model which coordinates activities in the neuromotor model. In the neuromotor model of handwritten stroke generation, stroke velocities are expressed as a Fourier-style decomposition of oscillatory neural activities. The timing network, which represents the timing action of BG, controls the events of the in the neuromotor model. The model gives a precise theory of what is loosely termed as motor preparation, involving dynamic interaction between BG and SMA. The model is further extended for multiple stroke production. The special emphasis given to BG in the models qualifies it as a candidate model for Parkinsonian handwriting. It is shown that model pathologies can capture several features of Parkinsonian handwriting like micrographia, irregular velocity profiles etc.