The Internet of Things (IoT) constantly offers new opportunities and features to monitor and analyze driver behaviour through wide use of smartphones, effective data collection and Big Data analysis. In that framework, the aim of this paper is to investigate the impact of mobile phone use on driving behaviour and road safety through the investigation of driving analytics collected by smartphone sensors. For this purpose, a 100-driver naturalistic experiment was carried out and an innovative data collection scheme using a smartphone application was exploited in order to record the respective driving performance data. Then mixed binary logistic regression models were developed in order to investigate whether mobile phone use during a trip is correlated with driving metrics and can therefore be accurately forecasted based on them. Finally, a model for all trips was developed, as well as models for trips on different road types (urban, rural, highway). Exposure metrics found to be significantly associated with the probability of mobile phone use are total trip distance and driving on workdays and during rush hours. Additionally, the average speed is negatively associated with the probability of mobile phone use while driving.