The reliability of normalized EMG(rms), (root mean square value of the electromyographic signal) measurements from musculus trapezius were evaluated for six variables in a field study: Static and median value of ADF (amplitude distribution function) and the time and number of periods when the load was below 1 and 2% MVC (Maximum Voluntary Contraction). The reproducibility of the method showed no significant changes between two measurements recorded from 12 VDU (Visual Display Unit) operators and the standardized agreement index (e) was negative only for the static value of the ADF. e = 1 -EI, where EI (agreement index) is the ratio between twice the standard deviation of the difference and the mean of the two measurements, The stability of the method in terms of stability ratio SDmm/SDv, was less than 1 for all variables, indicating a good stability. SDmm is the standard deviation of the difference between minimum and maximum value for each of the subjects for the six measurements. SDv is the average of the standard deviations of the six measurements at each visit, The reproducibility of the results was investigated by comparing one measurement from two groups, each consisting of 24 VDU operators doing similar type of work. No significant difference were found between the groups with regard to any of the six variables (0.07 <p < 0.87). In three studies, the relationship between trapezius load and the incidence of musculoskeletal illness in the neck and shoulder were investigated, Both the static load and the number of periods per minute when the load was below 1% MVC were found to relate to the incidence of such illness. However, the correlation between such illness and trapezius EMG parameters is relatively low, indicating a multifactorial etiology of musculoskeletal illness.