149 Transportation Research Record: Journal of the Transportation Research Board, No. 2484, Transportation Research Board, Washington, D.C., 2015, pp. 149–158. DOI: 10.3141/2484-16 High-occupancy vehicle (HOV) facilities have been widely implemented in the United States as a way to improve mobility, trip time reliability, and air quality of freeway systems. In California there are two types of HOV facilities, namely, limited access and continuous access with part-time or full-time operation. This paper presents a framework for comparing the mobility performance of the types of HOV facilities on the basis of regression analyses using a California data set as a case study. Parametric regression models were developed to explain empirical free- way capacity as a function of a set of explanatory variables including the type of HOV access. Modeling results consistently indicated that freeway segments with limited-access HOV lanes would have had higher overall capacities than those of continuous-access HOV facilities, every- thing else being equal. In addition, the proposed framework can be applied well to the analyses on other types of managed lanes. Since their introduction in the United States more than 40 years ago, high-occupancy vehicle (HOV) facilities have been integrated as a part of roadway network systems to relieve pressures from ever-increasing travel demands. It is generally accepted that HOV facilities can increase the passenger-carrying capability of roadway networks by providing incentives for travelers to carpool or use transit services. Such a traveling mode shift may essentially help metropolitan areas address their needs for improved mobility and productivity, while also being sensitive to environmental issues and quality of life (1). The planning, design, and operation of HOV facilities in the United States vary by state and sometimes by region. For example, in California the design and operation of an HOV facility are closely linked to the traffic demands along that freeway corridor, the geo- graphic distribution development, and the associated travel patterns in the region (2). In those areas that experience regular periods of congestion for many hours of the day, full-time HOV facility opera- tions with limited access are favored to maximize opportunities for increasing use and reducing travel time along HOV facilities. In other areas, however, where commute patterns generally consist of short definable peak periods and clear directional flows, part-time HOV operations are preferred during peak hours. With part-time opera- tion, the HOV lanes ideally should resemble general purpose lanes to minimize the potential confusion to motorists when they are open to general purpose traffic. Accordingly, it is preferable that access into and out of HOV facilities that operate part time should be continuous instead of limited (3). Traditionally, HOV facilities in Northern Cali- fornia have been operated with continuous access (Figure 1a) and at peak hours only, while those in Southern California have mostly been operated 24 h/day with limited access (Figure 1b) with a few exceptions. Compared with any other states, California has the most extensive network of HOV facilities in the United States, approximately 40% of the total HOV lane miles. As shown in Figure 2, there were about 1,370 lane miles of HOV facilities in California as of March 2013, and another 750 lane miles were under construction, programmed, or proposed. In essence, HOV facilities have been and will continue to be an integral part of California freeway systems. Therefore, it is important to ensure that these HOV facilities are being operated effectively and meet their purpose of improving mobility, trip time reliability, and air quality. To date, few studies have been conducted to examine specifically the differences in mobility performance between HOV facilities with different types of access control, which is the topic that motivates this research. With California as a case study, the objective of this paper is to present a framework for comparing the mobility performance of HOV facilities with different access types while accounting for other influencing factors, such as the number of lanes. The framework is based on conducting regression analyses on the dependence of free- way capacity (i.e., the response variable) on a set of explanatory vari- ables including the type of HOV access control (continuous versus limited access). The regression coefficient of this variable will then reveal the trend (positive or negative) and the magnitude of the dif- ference in mobility performance of one type of HOV facility over the other. This information is critical to the planning, design, and opera- tion of HOV facilities. It is imperative to point out that the regression models developed in this study do not mean to predict freeway capac- ity as is typically done in the Highway Capacity Manual 2010 (HCM 2010), but rather to serve as a tool for determining the statistical effect of HOV access type on freeway capacity (4). In addition, as stated in HCM 2010, some of the TRB methods may not be appropriate for freeway segments with HOV facilities. The rest of this paper is organized as follows. The next sec- tion reviews related studies on HOV facilities, especially those in California. Then data sources and data processing for the regression Different Types of High-Occupancy Vehicle Access Control Comparative Analysis of Empirical Capacities Guoyuan Wu, Kanok Boriboonsomsin, Matthew Barth, and Ramakrishna Tadi G. Wu, K. Boriboonsomsin, and M. Barth, Center for Environmental Research and Technology, University of California at Riverside, 1084 Columbia Avenue, Riverside, CA 92507. R. Tadi, California Department of Transportation, 464 West Fourth Street, 6th Floor, MS 619, San Bernardino, CA 92401. Corresponding author: G. Wu, gywu@cert.ucr.edu.