It is common that a survey aims to collect responses from specific population groups such as people with certain interests, people living in a certain area, people affected by a certain event, etc. In general, this kind of survey uses a registry or commercially available database to reach the targeted population in an effective and efficient manner. However, it is a challenging task as the registry or database information is often limited and expensive. In addition, primary available contact information is a phone number, which also limits the mode of survey to phone calls. During the last decade, web survey has become much more common as telephone surveys have experienced many challenges. In this study, a web-based survey with a multi-method sampling approach was adopted to conduct a motorcyclist survey in Florida. Motorcycles represent less than 3 percent of registered motor vehicles in Florida. The survey link was distributed through the following methods: Address-Based Sampling (ABS):Postcards with an invitation to complete the web-based survey were sent to a random sample of endorsed motorcyclists. In total, 30,000 addresses were sampled from the Florida DHSMV database of endorsed motorcyclists. Each sampled address was sent two invitation postcards over a period of two weeks. Facebook: An advertisement that targeted motorcycle riders ages 18 and older was placed on Facebook for 10 days to advertise the web survey. To increase participation, incentives were offered through drawings. Participants could enter to win a motorcycle safety-related prize, including a motorcycle jacket or a motorcycle helmet. In total,3,817 survey responses (including partially-completed surveys) were collected. The study explored the cost-effectiveness of Facebook advertisement compared to ABS in terms of reaching the targeted population and discussing the potential coverage errors by age.