This article describes a laboratory study of work preferences (ideal job demand and discretion levels) as moderators of the effects of paced and unpaced work on cognitive and affective responses. Posttest measures of cognitive performance and self-reported stress and arousal were used as outcome measures with covariance control for the corresponding pretest values. The experimental design allowed within-subjects contrasts of fast versus slow pacing and of machine-paced versus self-paced conditions. Self-paced performance compared favorably with machine-paced performance; however, individual differences in ideal demand influenced the relative speed of work under the two conditions. Work preferences also moderated relationships between pacing and outcome measures; ideal discretion moderated machine-pacing versus self-pacing effects in relation to cognitive performance and stress, and ideal demand moderated fast versus slow pacing effects in relation to arousal. These findings are discussed in relation to existing literature on pacing and on person-environment fit.