A parking information system consisting of a parking information website and roadside displays has been proposed for New York City to alleviate parking difficulties. To justify the implementation of such a system, it is necessary to investigate parking search time and its relationship with parking information. In addition, the parking information items needed for the website and roadside displays and the target markets for this information must be identified. In the investigation, a survey was conducted to collect data on parkers' demographic and trip characteristics. On the basis of the collected data, the effect of parking information on parking search time was studied, and an ordered probit model was developed. To identify parking information needs, the analysis of variance method was employed to determine the rank of the information items for the website and roadside displays. The profiles of the parkers were also identified for each information item using developed models. The results of the investigation indicated that parking search time is significantly influenced by the amount of parking information parkers possess before starting their trips. Six information items were identified to be the most necessary for the parking information website, and five as most necessary for the parking roadside display. All these information items should be considered when designing a parking information system.