This paper represents a preliminary effort in considering how protein motors could harness thermal fluctuations to generate force and movement. The initial premise for this model is the thermal motor described by Feynman which consists of a ratchet and an interdigitating, spring-loaded pawl. By analogy, one can imagine that biological motors interact weakly with their filament subunit substrate and that thermal fluctuations displace the motor to adjacent subunits on the filament. Unidirectional motion ensues if ATP energy either changes the energy of a spring-like component in the system or asymmetrically alters the energy barrier to displacement. Although a simple thermal ratchet model can account for the maximal forces and velocities produced by biological systems, it does not adequately explain the force produced and the energy expended by muscle as a function of its velocity of shortening. To explain these phenomena, we propose that the energy barrier of the thermal ratchet changes as a function of load. A load or velocity-dependency in the transition of the motor from a weak to a strong binding state could produce this effect. The thermal ratchet model for energy transduction can also explain many of the observations of filament translocation along motor-covered surfaces in the in vitro motility assay. Furthermore, unlike the rotating cross-bridge model which predicts a large conformational change in the motor, unidirectional motion and force production with a thermal ratchet motor could be accomplished through small structural alterations in the motor head and the filament subunit. In general, models are useful if they formulate a set of predictions that serve as guides for future experimentation. First of all, it should be instructive to examine more critically the contact between motors and filaments and to determine the quantal episodes of displacement and force. Furthermore, it will be important to inspect the filament, in addition to the motor, for conformational changes that occur throughout the ATPase cycle. Along this line, the ATP hydrolysis reaction should be reexamined to determine if a large number of energy states of the motor protein and the filament occur during the hydrolysis cycle. Furthermore, our model makes predictions regarding the relationships between the energy barrier height (related perhaps to motor-filament binding affinity) and velocity or force. Whether the weak to strong binding state transition is dependent upon load or velocity is another unique prediction of our model that could be experimentally probed.(ABSTRACT TRUNCATED AT 400 WORDS)