The objective was to obtain research-based, holistic knowledge about necessity and effect of practiced measures against L. pneumophila in municipal shower systems in Stavanger, Norway. The effects of hot water treatment and membrane-filtering were investigated and compared to no intervention at all. The studies were done under real-world conditions. Additionally, a surveillance pilot study of municipal showers in Stavanger was performed. The validity of high total plate count (TPC) as an indication of L. pneumophila was evaluated. A simplified method, named "dripping method", for detection and quantification of L. pneumophila was developed. The sensitivity of the dripping method is 5 colony-forming units of L. pneumophila/ml. The transference of L. pneumophila from shower water to aerosols was studied. Interviews and observational studies among the stakeholders were done in order to identify patterns of communication and behavior in a Legionella risk perspective. No substantial effects of the measures against L. pneumophila were demonstrated, except for a distally placed membrane filter. No significant positive correlation between TPC and L. pneumophila concentrations were found. L. pneumophila serogroup 2-14 was demonstrated in 21% of the 29 buildings tested in the surveillance pilot. Relatively few cells of L. pneumophila were transferred from shower water to aerosols. Anxiety appeared as the major driving force in the risk governance of Legionella. In conclusion, the risk of acquiring Legionnaires' disease from municipal shower systems is evaluated as low and uncertain. By eliminating ineffective approaches, targeted Legionella risk governance can be practiced. Risk management by surveillance is evaluated as appropriate.