This study investigated the comparability of paper and computer versions of a functional writing assessment administered to adults 16 and older. Three writing tasks were administered in both paper and computer modes to volunteers in the field test of an assessment of adult literacy in 2008. One set of analyses examined mode effects on scoring by comparing scores for adults' original handwritten responses to scores for their transcribed word-processed versions. Differences in overall or individual criterion scores were either statistically or practically nonsignificant. A second analysis examined differences in performance between adults who received paper versus computer versions of the same task. A third analysis examined whether the differences between administration modes were moderated by demographic variables or by computer experience. Results showed that adults performed better overall and on most aspects of the writing tasks when writing on paper than on computer. The effects of administration mode varied by employment status for one task, by race/ethnicity, age and word-processor experience for another task, and remained the same for all subgroups for the third task. These results suggest that, depending upon the writing tasks, a computer mode of administration may disadvantage some subgroups (e.g., unemployed) more than others (e.g., employed). (Contains 13 tables.)