As part of a continuing effort to improve existing methodology for development and use of occupational data gathering devices, this report considers the relative validity of two varied item formats for obtaining length of service data; specification in total months versus a combined years and months formulation. The later format was found to have much greater validity for relatively lengthy time periods but no significant difference was found for shorter periods. An alternative method for obtaining length of service data which employs a computer determination of elapsed time based on respondent provided dates is recommended for operational use. (Author)