Karasek's job demands–control (JDC) model posits that job control can buffer against the harmful effects of demands experienced by employees. A large volume of JDC research has obtained support for the main effects of demands and control, but not the interactive effects. Recent research on the challenge–hindrance stressors framework, however, found that work stressors may not always be deleterious, suggesting alternative hypotheses about the effects of demands and control. The present study therefore examined competing hypotheses concerning the effects of job demands on occupational health outcomes. Using a sample of 316 employees in a Chinese manufacturing company, we found that, consistent with the challenge–hindrance framework, production demands were challenge stressors associated with favourable outcomes (i.e. job satisfaction and psychological well-being). In addition, results showed that the interactive role of job control depended on the nature of outcome variables. Future recommendations and implications of findings are discussed. Copyright © 2014 John Wiley & Sons, Ltd.