In this study it was possible to prove the Werther effect in suicides after watching fictional models for the first time. A twice-broadcast (1981, 1982) six-episode weekly serial showing the railway suicide of a 19-year-old male student provided a quasi-experimental ABABA design to investigate differential effects of suicide imitation. Imitation effects were most clearly observable in the groups whose age and sex were closet to those of the model. Over extended periods (up to 70 days after the first episode), the number of railway suicides increased most sharply among 15- to 19-year-old males (up to 175%); the effect steadily decreased in the older age groups, so that no effect was observable for males over 40 years and females over 30 years. Also, the imitation effects remained detectable for longer periods in the groups closest in age to the model. The increases observed after the first and second broadcast for males aged less than 30 years closely corresponded with the respective audience figures for the two showings.