A clinical rating scale which measured the severity of tremor in 20 patients (12 with essential tremor and 8 with "dystonic" tremor) was assessed at specific anatomical sites for both inter and intra-rater reliability using four raters. The scores obtained with the scale were compared with the results of upper limb accelerometry, an activity of daily living self-questionnaire and estimates of the tremor induced impairment in writing and drawing specimens. The results show that, for the purposes of routine assessment and therapeutic trials, a clinical rating scale can produce reliable results which are a more valid index of tremor induced disability than standard postural accelerometry.