This study evaluated the potential positive effects of stressors on job performance by examining the shape of the relation between stressors and job performance. The 281 respondents were a demographically and organizationally heterogeneous group from the Detroit area, who were employed during the study. They were given four structured in-home interviews, approximately 6 weeks apart, over a period of 18 weeks. Interviews were also conducted with a significant other, nominated by each respondent from work life. Stressors examined were role ambiguity, role conflict, and job insecurity. Strains, also examined as potential stressors, were job dissatisfaction, anxiety, depression, and anger. Technical and social aspects of respondents' job performance were measured separately, as were absenteeism and tardiness. All zero-order Pearson correlations were either statistically significant and in predicted directions or essentially zero. All relationships were monotonic, suggesting that, for these stressors, their optimal amounts are generally zero. Results are discussed in terms of arousal and activation, information-processing, and expectancy theory.