Transcranial magnetic stimulation (TMS) was compared to proton magnetic resonance spectroscopy (1H-MRS) for the detection of upper motor neuron loss or dysfunction in 49 ALS patients classified according to the El Escorial criteria. Abnormal NAA/Cho ratios were detected in 53% of ALS patients. Abnormal TMS results (i.e. cortical inexcitability or prolonged CMCT's) were obtained in 63% of ALS patients. If one or both methods were considered for diagnosis of upper motor neuron degeneration/dysfunction, the percentage of abnormal findings was 77%, whilst in 39% of all patients both methods produced abnormal results. Compared to TMS, 1H-MRS detected more patients with upper motor neuron involvement in the suspected El Escorial subgroup (42% versus 25%), whereas TMS detected more patients with upper motor neuron involvement in the possible (81% versus 50%), probable (71% versus 57%) and definite El Escorial subgroup (71% versus 64%). We conclude that the combined use of 1H-MRS and TMS increases diagnostic accuracy for the detection of upper motor neuron involvement in ALS patients.