Finger-to-nose test, fine finger movements, to maintain arms against gravity, alternate movements of hands, Barany's test, muscle tone evaluation, Stewart-Holmes test and handwriting are the conventional clinical tests most frequently used in daily practice to evaluate voluntary movements of the upper limb. We describe a new clinical manoeuvre consisting of a horizontal pointing movement of one upper limb (the moving limb) towards the contralateral motionless limb. We have analyzed the reliability of our new test and of the conventional tests between 2 observers using kappa statistics in a group of 34 right-handed neurological patients. Agreement is very good for our test, since k has a value of 0.63 for left upper limb and a value of 0.64 for right upper limb. The only conventional test characterized by a higher interobserver reliability is handwriting. Furthermore, our manoeuvre is original, as attested by partial association coefficients analysis between our test and conventional tests. This might be due to the fact that our manoeuvre is the sole test investigating rapid proximal movements towards a fixed area in space. In conclusion, our horizontal pointing manoeuvre has a very good reliability between 2 observers and appears original.