Source monitoring encompasses memory (of item and source) and reconstructive (guessing) processes while determining the episodic context of previous experiences (Johnson et al., 1993). Based on observable response frequencies, the two-high-threshold multinomial processing tree (MPT) model of source monitoring (2HTSM; Bayen et al., 1996) provides probability estimates of these memory- and guessing-based processes. Heck and Erdfelder (2016) proposed a formal modeling approach integrating response times (RTs) into MPT models to estimate the relative speed of latent processes. Here, we present an application of this MPT-RT rationale to source monitoring and introduce the RT-extended 2HTSM model (2HTSM-RT). We conducted a standard source-monitoring experiment with words as items and screen positions as sources (N = 35). After categorizing continuous RT data into discrete bins from fast to slow responses separately for each individual, we calculated frequencies for all combinations of discrete attribution responses and RT bins. The main objective of the current study is to test whether the 2HTSM-RT fits RT data of a typical source-monitoring paradigm and to present more holistic information about the processes involved in source monitoring by understanding the relative speed of processes (reflected in the latency parameters) in addition to the accuracy of the process outcomes (reflected in the standard 2HTSM model parameters).