Each of the sleep and wake states of animals are characterized by specific patterns of behavioral, motoric, and electrophysiological activity. Sleep-wake behavior of rats, rabbits, and human infants (3 of each species) was observed, and, at the same time, a single-channel analog recording was obtained of the motoric activity. A judge who had extensive experience in observing sleep-wake behaviors of human infants, but who was unfamiliar with these behaviors in nonhuman species, scored the analog signals of the rats and rabbits. Another judge, who knew rat and rabbit state behaviors well, but who had not had experience observing human infants, judged the states of the human infants from the analog signals. These judgments were evaluated by comparing them with the findings obtained from direct observation of the subjects. For the nine subjects, the correlations between the judges' scoring of the analog recording and direct observation ranged from 0.856 to 0.985. A more stringent criterion is to determine the exact agreement between judge and observer for each 10-s epoch during the observation. For the major states of active sleep and quiet sleep, the exact agreement ranged from 95.2 to 100%. The wake state was easily judged for rats (100% agreement) but was more difficult for rabbits and humans (59.9 and 66.7%, respectively). This ability to accurately score the state behavior of an unfamiliar species from an analog movement record alone represents a cross-species invariance in neuromotor organization. The critical parameters appear to be regularity or irregularity of respiration and the temporal distribution of gross motor movement.