Studies have suggested that sitting time at work may lead to underperformance but they may underestimate the benefits to desk workers' performance of reducing occupational sitting time without considering the relative effects of the specific activities replaced.To estimate differences in work performance (presenteeism, absenteeism and engagement) when occupational sitting time is reallocated to standing/stepping in desk workers.Data for middle-aged desk workers were from a Japan-wide online survey (n = 2228). Self-report proportion of occupational sitting and standing/stepping, work hours and work performance indicators, including absolute (ratings relating only to self) and relative (ratings of self, compared to others) presenteeism and absenteeism, and dimensions of work engagement, were collected. Partition and isotemporal substitution models were used to investigate the associations of occupational sitting and standing/stepping time with work performance, including their reallocation effects.In partition models, longer occupational sitting time was associated with a lower absolute presenteeism score (i.e. less productivity), lower absolute absenteeism (i.e. longer-than-expected work hours), and lower engagement. Longer occupational standing/stepping time was associated with lower absolute absenteeism and more engagement. Isotemporal substitution models showed that each hour of occupational sitting reallocated to standing/stepping was favourably associated with overall work engagement (B = 0.087; 95% confidence interval 0.051, 0.122) and its dimensions (B ranged from 0.078 to 0.092), but was not associated with presenteeism or absenteeism.These findings suggest that management support and practical initiatives to encourage desk workers to replace portions of their sitting time with standing/stepping may contribute to enhanced work engagement.