One limitation of hired farm labor data from the Decennial Census is undercounting; the seasonal nature of farm work means that many farm laborers are not employed in March, when the Census is taken. The 1981 Hired Farm Working Force Survey provided data to evaluate the usefulness of Census data for farm labor research. Data were grouped into farmworkers employed during March (representing the hired farmworker population measured by the Census) and farmworkers employed at other times of year and therefore excluded from Census farm labor categories. Step-wise discriminant analysis was used to determine whether the groups differed significantly by demographic, employment, educational, and economic characteristics. Of 20 variables used, 15 were statistically significant discriminators between the groups, with farm earnings and hired farmwork as the major activity during the year being the two most significant. March farmworkers were generally those committed to and economically dependent on hired farmwork for most of the year. However, almost one-third cited nonfarm work or not in the labor force as their primary activity during the year; they would not identify with farmwork as an occupation. About 9% of the other month workers cited hired farmwork as their major activity, yet would not be included in Census data as farmworkers. (MH)