AbstractHand‐dynamometer induced muscle tension was manipulated in the context of an experimental paradigm that included tests of discrimination and generalization with rigorously defined schizophrenic and affective patients, as well as schizotypic and normal controls (totalN= 44). The experiment was designed to test predictions that derive from Broen and Storms' (1967) learning/motivational analysis of schizophrenic performance deficits. The results failed to support the theoretical predictions, but did reveal that the ability to maintain induced muscle tension within prescribed limits discriminated reliably between the groups (p⩽.03) and that this discriminability increased with increasing tension demands (p⩽.001). The data are discussed in terms of possible relations to basic sensory/motor (i.e., neuromuscular and/or proprioceptive) mechanisms as well as to more central psychological (i.e., attentional/cognitive) mechanisms, and directions for future research are out