Background: We estimated the multiple sclerosis (MS) incidence in the Netherlands for better active monitoring of potential vaccine safety signals. Methods: A retrospective cohort study (1996–2008) was conducted using a population-based general practice research database containing electronic medical records. Additional information was collected to validate incident probable cases. Results: In the source population (648,656 persons), 146 incident probable MS cases were identified. Overall incidence rate was 6.3/100,000 person years (py; 95% CI, 5.2–7.2). In the subgroup in which MS could be fully validated, the incidence increased from 4/100,000 py (95% CI, 3–5) in 1996–2004 to 9/100,000 py in 2007/8 (95% CI, 6–16). This increase was highest among women, but not statistically significantly different by gender. The median lag time between first recorded symptoms and MS diagnosis decreased from 32 months (<1998) to 2 months (>2005). Conclusions: MS is rare in the Netherlands. In recent years, there was a slight increase in the incidence especially among women during the fertile age. This increase coincided with a decrease in lag time between symptoms and diagnosis, both for men and women. This trend should be taken into account in the interpretation of MS cases occurring in a population where new vaccinations will be introduced shortly.