Communicated by W. SprößigIn this article, we propose a mathematical model that describes the dynamics of a population divided into susceptible drinkers, moderate drinkers, and heavy drinkers subject to an external influence. The external influence is modelled using a supplementary dynamical variable that is not a group of individuals but that enters the equations affecting the choices of the population classes. The system we define can be investigated using two simplified systems (one of which is a real subsystem), which model the populations of susceptible and moderate drinkers or susceptible and heavy drinkers independently. The dynamics of these two subsystems can be described exhaustively. The full system is too rich in possible scenarios, but its qualitative behaviour is connected to that of the two simplified systems. We make a complete description only in one particular case by means of numerical simulations.Alcohol is well known to cause a multitude of negative effects such as change of mood, behaviour, and coordination, heart problems, liver problems, and cancer [1]. In recent times, particularly in Western society, peer pressure drives youngsters towards an extreme behaviour denoted as binge drinking. The concept of binge drinking is defined, for instance, in [2], and it consists in drinking large amounts of alcohol to become heavily intoxicated. Binge drinking is primarily a practice of youths in the 17-30 age group, and people that indulge in binge drinking are likely to exhibit anti-social behaviour. To binge drinking can be attributed losses of the order of billions of dollars in the USA alone, caused by diminished productivity and health issues associated with auto-related accidents, traumas, and crimes (including physical and sexual assault) [3]. Binge drinking can affect well over 25 per cent of the population as may be seen from specific data for Durham City [4] and Stockton-on-Tees [5] in North East England, with some areas being as high as 28.93%, in, for example, Newcastle-upon-Tyne, [4,5]. In addition, the death rate due to drinking alcohol has risen dramatically in the UK with, for example, Glasgow recording a death rate of 83.7 per 100,000 for both men and women during 1988-2004 [6].Drinking we believe has been investigated using the techniques of epidemical models, and such approach has proven to accurately predict the percentages of binge and moderate drinkers [7][8][9][10][11][12][13][14].What has never been modelled in such a field is the influence of external variables that favour or discourage such habits. This external influence can be modelled introducing the concept of an information variable. To our knowledge, the idea of expanding an epidemiological system introducing an information variable can be found for the first time in [15][16][17][18][19][20]. In such works, the authors investigate a vaccinating behaviour where the information variable measures the publicly available information on the state of the disease. This influences the adhesion to vaccination programmes which, i...