Objective: To examine the associations between sickness presenteeism (SP) reported as number of days with SP reported as number of times and to evaluate their responsiveness.

Methods: The study population (n = 454) consisted of employed individuals, at risk of long-term sickness absence. Correlation analyses were performed to examine associations between the two SP measures and external constructs such as work performance, general health, and registered sick leave. Both SP constructs were measured several times to examine responsiveness.

Results: The SP measures are moderately correlated. They moderately correlated with work performance and health status measures. SP reported as number of times seems to be more sensitive than number of days in detecting changes after rehabilitation.

Conclusions: Numerical or categorical constructs are valid sources of data on SP. However, categorized SP seems to be more responsive.