The present-day view of the neural basis for the senses of muscle force and heaviness is that they are generated centrally, within the brain, from copies of motor commands. A corollary of the motor discharge generates a sense of effort which underlies these sensations. In recent experiments on force and heaviness sensations using thumb flexor muscles, a rather different explanation has been invoked: Subjects were proposed to rely predominantly on inputs of a peripheral origin, in particular, the signals of muscle spindles. The present experiments have been carried out at the elbow joint to determine whether these new ideas apply more widely. The effects of fatigue of elbow flexor muscles have been studied in force and heaviness matching tasks using three exercise regimes, a sustained maximum voluntary contraction (MVC), a maintained contraction of 35 % MVC, and a maintained contraction of 35 % MVC combined with muscle vibration at 80 Hz. In force-matching experiments, subjects were required to contract both arms and while the reference arm generated the target force under visual control, it was matched by the indicator arm without visual feedback. During the 100 % MVC exercise, force in the exercising reference arm fell rapidly to almost a half of its original value over 90 s while force in the indicator did not fall, leading to a significant overestimation of the reference force. During the 35 % MVC exercise, subjects also overestimated the reference force and this persisted at 5 and 10 min after the exercise. When 35 % MVC was combined with vibration, the amount by which the indicator arm overestimated the reference force was significantly reduced. In heaviness matching experiments, subjects could move their arms through a small range. The reference arm was loaded with a weight, and weights were added or removed from the indicator until heaviness felt the same in the two arms. There was a small, but significant fall in the matching weight used after 100 % MVC exercise, that is, the weight held by the fatigued arm felt lighter. The 35 % exercise did not alter heaviness sensation while 35 % MVC exercise with vibration led to a significant reduction in perceived heaviness. To conclude, while the results of these experiments on elbow flexors are not as clear cut as for thumb flexors, the central effort hypothesis falls short, in a number of respects in explaining the data which are able to be interpreted in terms of a peripheral afferent contribution to the senses of force and heaviness.