Animal (monkey, rat, mouse) models are widely used to investigate degenerative processes and potential therapeutic treatments for human Parkinson's disease (PD). One task that has proved useful in these investigations is a reach-to-grasp task (skilled reaching) in which an animal reaches for a piece of food that it then consumes. Rats with extensive unilateral Dopamine depletions are impaired in using the contralateral limb. The qualitative features of posture, lifting and advancing the limb, pronating the paw to grasp food, and in withdrawing and supinating the paw to place the food in the mouth are impaired, as is reaching success. Humans with PD are often described as having poor manual dexterity that worsens as the disease progresses. As there have been no detailed comparisons of reaching movements in the animal models and in PD subjects, the following descriptive analysis was performed. Ten subjects with PD, eight age matched controls and 14 young normal subjects were studied as they used a natural movement of reaching for a small piece of food that they then placed in the mouth to eat. The reaching movements were described using Eshkol-Wachman Movement Notation (EWMN), supplemented with kinematic analyses. From this description, a 21-point rating scale was devised to describe the component movements of the reach. Movements included: orienting the head and eyes to the target, adjusting posture, lifting the hand, shaping and aiming the digits to the target, pronating the hand to grasping the food with a pincer grip, lifting and supinating the hand to transporting the food to the mouth, and further supinating the hand and opening the digits to place food in the mouth, and finally returning the hand to the starting position. Analysis indicated that most aspects of the reaching movements of the PD subjects were significantly different relative to both young control subjects and old control subjects. As compared to the control groups, postural and reaching components of the movements were fragmented, movements were achieved using more proximal segments of the body, and rotatory movements of the hand were limited. The PD subjects did use a pincer grasp to obtain the food, but the grasp was less independent of other digit movements than was observed in the control subjects. These results are discussed in terms of a homology to impairments displayed animal models of PD.