Effects on driver capabilities
According to an overview of studies carried out in laboratories, driving simulators and instrumented vehicles, Muscovite and Robinson  come to the conclusion that most skills related to the driving task already start to deteriorate at a BAC-level as low as 0.2 g/l.
The driving task can be divided in three subtasks. The first group of subtasks includes the tasks on the operational level. These are the actions that have to be carried out to keep speed and course. They include steering, changing gear, accelerating, braking but also other manual and mostly fully automated actions for manoeuvring and keeping the vehicle in an optimum operational state (i.e. switching on the windscreen wipers) while driving. The second group of subtasks includes the tasks at the tactical level. These are the decisions one has to take when participating in traffic. This consists of the application of the rules of the road (i.e. I have to yield for that other car) and decisions concerning manoeuvres that include other road users (i.e. now I can safely overtake that other car). The third group includes the tasks at the strategic level. These tasks deal with vehicle choice and route choice. In this particular case one can think of the decision a driver has to take whether he will drive or not after having consumed alcohol.
Alcohol affects task performance at all three levels. However the overwhelming majority of the research that has been carried out is on the effects alcohol has on the tasks at the operational level and the tactical level. Caird et al  give a recent overview of these studies. For the operational level of the driving task, their conclusions are:
- Tracking performance (keeping course) starts to deteriorate at a BAC as low as 0.18 g/l. Reductions in performance with respect to keeping a constant distance behind a leading vehicle (keeping headway) starts at a BAC of 0.54 g/l when the leading vehicle keeps a constant speed. When the leading vehicle changes speed, reductions in performance start at a BAC as low as 0.3 g/l
- Reaction times when driving get longer. There is a difference between a driver's capability on simple reaction time tasks and choice reaction time task. In a simple reaction time task a driver has to press a key as quickly as possible after a stimulus (auditory or visual) has been presented. In a choice reaction time task a driver has to respond differently to two stimuli by pressing one key for event A and a separate key for event B. Choice reaction time begin to deteriorate at a BAC of 0.6 g/l, and for simple reaction time task the BAC is considerably higher before significant prolonged reaction times appear
- Reactions on a visual detection task (perception) when driving starts to decrease significantly at a BAC of 0.8 g/l.
At the tactical level:
- Decrease in the ability to divide attention between the driving task and another task starts at BACs between 0.3 and 1 g/l (depending on the complexity of the second task). When drivers have to divide their attention between driving and another task (i.e. having a conversation with a passenger) and this ability starts to deteriorate because of alcohol, subjects tend to focus on one of the two tasks at the expense of the other
- When BAC increases, drivers tend to fix their eyes more on the central visual field and fewer eye movements are made to the peripheral view. When under the influence of alcohol, drivers use fewer sources in the visual field to obtain information about the environment, they take longer to recognize and respond to aspects that present vital information about their environment (i.e. street signs) and they focus their attention on aspects occurring in their central field of vision often at the cost of peripheral information
- The increase in the number of mistakes and prolonged reaction times when drivers are confronted with a complex secondary task, even when small quantities of alcohol are consumed, indicates that alcohol causes information processing to be hampered.
The impact of alcohol on the performance of a driver at the strategic level cannot be studied in driving simulators or instrumented vehicles. However, according to the Theory of Planned Behavior (TPB) , alcohol must have a significant impact on the strategic level. The TPB states that intentions are influenced by three mechanisms: attitudes, subjective norms and perceived behavioral control (pbc). Attitudes towards certain behavior reflect the degree of positive or negative evaluation the individual has towards performing it (i.e. drink driving is dangerous). Subjective norms refer to the perceived social pressure to engage or not engage in certain behavior. This reflects what one's 'important others' would think about the intended behavior (i.e. my friend would disapprove when I drive while I'm drunk). PBC reflects the perceived ease or difficulty of undertaking a given behavior (i.e. if I want to, I could easily drive safely when I'm drunk). Alcohol consumption leads to loss of self-control and thus it has an effect on PBC. After having consumed alcohol a driver is much more inclined to think that he can easily drive safely when he is a bit drunk.