People with intellectual disabilities (ID) have been overrepresented in the suspect and defendant populations. However, the two key language-related procedural moments which affect their legal status (police caution and courtroom communication) have been traditionally studied separately. Objectives: To compile and review empirical evidence from the peer-reviewed literature both at the investigative and trial stages, determine if the linguistic nature of cautions is associated with poorer understanding, and propose an agenda for future research. A targeted integrative review of 35 peer reviewed sources and an exploratory analysis of the readability and linguistic features of caution wordings in standard format. Even in optimal conditions (when the caution is sent by sentence and not read at once) there is an extremely low rate of full comprehension (10% of suspects, 96% reporting comprehension). In recent experimental studies with adults with ID (n = 30), none of the participants had a sufficient understanding of the Scottish caution regardless of listenability modifications. The present alert, which contains a legally complicated adverse inference warning, has a Flesch-Kincaid Grade Level (FKGL) of 5.2, a result that is easier to understand than the previous (pre-1995) alert, which had an FKGL of 8.0. This means that the surface readability measures do not accurately predict comprehension. Other adults with mild developmental disabilities had 75% correct for only 8 of 34 legal terms (vs. 28 of 34 for non-delayed controls at an 85% criterion). In addition, procedural support for witnesses with ID continues to be poorly developed in the courtroom, and former defendants report very little procedural support. The Cumulative Procedural Barrier Model proposes that when non-comprehension goes unrecognized during the caution stage, it will be systemically carried forward in police interviewing, plea negotiation, and trial. Some future research directions of particular interest are: developing and testing comprehension checks that are valid for students with ID, identifying and testing interventions that are targeted to ID students, and conducting courtroom studies of defendants which are linked-stage empirical designs and cross-jurisdictional studies.