Traditional measures of speed obtained through traffic observations are not based on detailed information about the related drivers and vehicles. Data from naturalistic studies, such as SHRP2 – NDS, can mitigate this issue by combining the key data on driver, roadway, and speed choice behavior. The objective of this study is to assess drivers’ speed choice on freeway ramps as a function of ramp design, trip summary, and driver characteristics. The data analysis provides insights into various spatial and temporal factors. To conduct the data analysis time series reduction, matching, and clustering methods were implemented to define a new speed choice behavior response variable denoted as driving state. Using the resulting response variable and the three groups of predictors, neural network analysis was conducted to identify the most influential predictors and their effects on the speed choice behavior of drivers during on-ramp and off-ramp travels. Results of this analysis of speed choice behavior on freeway ramps indicate that the speed choice at these locations is indeed a complex process and is mainly influenced by the temporal and traffic conditions. Personal characteristics of drivers were also found to influence speed choice in these locations.