Pedometer use and step count goals have become popular in physical activity (PA) interventions in different settings. Previous pedometer-based workplace interventions were short term, uncontrolled and executed outside Europe. This European quasi-experimental study evaluated the effects of a 20-week pedometer-based PA workplace intervention. Pedometer-based and self-reported PA from one intervention worksite (68 participants at follow-up) was compared with the data of a comparison workplace (79 participants at follow-up). A downward trend in overall step counts from baseline (end of summer) to follow-up (winter) was found (F = 3.3, P = 0.071). However, the intervention effect revealed a significant smaller decrease in the intervention workplace (-618 steps/day) than in the comparison workplace (-1389 steps/day) (F = 8.8, P = 0.004). This intervention effect was only present in already active participants, reaching 10 000 steps/day at baseline (intervention participants: -1706 steps/day; comparison participants: -4006 steps/day) (F = 5.5, P = 0.023). Overall project awareness was very high (97%) and the intervention strategies were judged "good to very good" by 57-95% of the participants. However, the proportion of intervention participants reporting that they had changed their PA behavior because of the intervention (31%) and reporting that they had used the pedometer during the intervention (48%) was limited. Future workplace projects should give extra attention to inactive employees. (Contains 3 tables and 1 figure.)