Although Web-based data collection has now been established, experimental mood research via the WWW is still in its infancy. If consumer psychology wants to study the influence of mood on consumers' behavior and attitudes towards online shops and services, we need reliable mood induction procedures that work in an online environment. This presentation illuminates the methodology of Web-based mood manipulation. It presents evidence for its feasibility and for its limits. Several online mood induction procedures were tested in three experiments. Experiment 1 examined the effectiveness of the Velten method and of mood-suggestive photographs to induce a positive and negative mood, each in a long and a short version. Long versions induced negative mood more effectively than short versions. However, neither version was successful in inducing a positive mood. In Experiment 2, cartoons and jokes were tested to induce a positive mood. Cartoons turned out to be effective, whereas jokes did not. Experiment 3 examined two methods in which participants were asked to bring up associations to affectivelyvalenced words. Both association tasks were not suitable to call forth a positive or negative mood.