Many in-vehicle systems have multimodal interfaces and allow mobile device connectivity to embedded and cloud-based systems. However, some drivers continue to use their hand-held devices regardless of whether the device can be synched with the vehicle. In these situations, drivers appear to rely more on voice interfaces given the perception that they may be safer to use for performing in-vehicle tasks. Drivers use of these systems within their own vehicles was explored using contextual interviews conducted in Seattle, WA and Rockville, MD. The study was conducted on planned routes with 64 participants, who were asked to perform voice control tasks for communication, navigation, accessing information, and entertainment while driving. The interviews were video recorded and a researcher rode along and logged the responses and errors made during the voice interactions. Cognitive workload was assessed using the NASA TLX (Task Load Index). The study showed that drivers are willing to use various voice control devices using smartphones, original equipment manufacturer (OEM) installed voice systems, or a combination. The types of errors made and the level of workload perceived varied among the different voice tasks, indicating that varying levels of distraction can occur with voice interfaces.