In the nineteen seventies, Steven Krashen put forward what has become both a controversial and an influential model, which he referred to as the Monitor Model (Krashen, 1979, 1982, 1994). He actually put forward five hypotheses, which are sometimes given the name of the Input Hypothesis (his fourth hypothesis) and sometimes the Monitor Model (his third hypothesis). His basic thesis was that adults can learn a language consciously from explicit transmission of declarative knowledge, but that this learned knowledge does not help them to communicate spontaneously. In order to achieve this, they need to acquire the language from exposure to comprehensible input (see Chapter 1 in this volume). In Krashen’s view, the only role for learned declarative knowledge was to help learners monitor their utterances prior to and subsequent to production of language. While acknowledging that this could be a useful (but limited) role, he pointed out that excessive monitoring could slow down a speaker and make their hesitant speech very difficult to listen to. He also categorised learners into three types: over-monitors, who nearly always monitor their production and suffer the communicative consequences of hesitation and reticence; under-monitors, who rarely monitor their production and thus make many errors; and optimal monitors, who monitor when it is appropriate and significant (e.g. in formal writing or in speech situations where accuracy is important). Whilst not disagreeing that the main role of explicitly learned knowledge is monitoring, we would go further than Krashen and say that such monitoring takes place before, during and after production, that monitoring is not just of accuracy but of fluency, appropriateness and effect, that implicitly acquired procedural knowledge can play a crucial role in spontaneous monitoring, that effective monitoring facilitates effective production and that effective monitoring can make a valuable contribution to language acquisition. Amazingly, given its importance in the process of producing language, monitoring has received very little attention in the research literature on language production, an area which is itself under-researched in our view. In this chapter we will refer to what research literature there is on monitoring and language production, but much of the chapter will be theoretical in the sense that we will be using our experience as researchers in a number of academic fields, our experience as language users and teachers and our informed introspection and intuition to posit a description of the mental processes involved. For some SLA researchers such theorising is taboo and only empirical evidence is acceptable. For us, theorising can help to fill the gaps in the research literature by stimulating thought, provoking controversy, inspiring application and suggesting profitable areas of research. After all, in applied linguistics there is an honourable tradition of philosophical theory with such influential figures as Pit Corder, Strevens, Chomsky, Halliday, Widdowson, Stevick, Faneslow and Maley inspiring thought and exploration with their novel theories rather than with their empirical evidence.