In case of a fall or crash, the use of a bicycle helmet was found to reduce serious head/brain injury[i] by 60% and fatal head/brain injury by 71% on average. These are so-called ‘best estimates’, meaning that it has been established with 95% certainty that the risk reduction for serious head/brain injury is between 54% and 65% and for fatal head/brain injury between 44% and 85%. This has become apparent from the most recent meta-analysis by Høye . This study compared injuries of helmeted and non-helmeted cyclists by combining the data of 55 (primarily) case-control studies from different countries meeting strict scientific requirements. This meta-analysis is a follow-up to a previous meta-analysis by Olivier & Creighton  that showed similar effects: a 69% reduction of the risk of serious head/brain injury (with 95% certainty of a reduction between 63% and 75%) and a 65% reduction of the risk of fatal head/brain injury (with 95% certainty of a reduction between 12 and 86%). Both studies show that, in general, the protective effect of bicycle helmets is the same for both children and adults  .
Most studies from the aforementioned meta-analyses have been carried out in the United States, Canada and Australia, and a few in Asia and Europe; none of these in the Netherlands however. In those countries, the infrastructure and traffic composition as well as bicycle use are often different from the Netherlands. It is possible, that bicycle crashes in the Netherlands may be of a different nature, which has consequences for the protective effect of bicycle helmets. However, there is no a priori indication of a larger or smaller effect in the Netherlands.
Bicycle helmet effectiveness is based on case-control studies, which is the prevailing method of study for this subject (see below for more information). In addition, biomechanical studies, studies involving computer simulations, and time series analyses have been done (also further explained below). In general, these studies find a considerable effect on reducing head/brain injury resulting from a fall or crash, although in time series analyses the effect is smaller.
In case-control studies, the effectiveness of bicycle helmets is established by comparing cyclists involved in crashes resulting in head/brain injuries (case) to injured cyclists without head/brain injuries (control). The major advantage of case-control studies is that they concern real crashes that have actually happened. Case-control research into the effectiveness of bicycle helmets is done when there are no data about the (differences in) bicycle kilometres travelled by helmet users and non-users (exposure). This is usually the case when cyclists with or without helmets are compared. This method has, however, also been criticised, since it could result in an overestimation of the effectiveness of bicycle helmets . Yet, other researchers did not find any evidence of this overestimation ( for example). In general, a well-designed case-control study is considered as a reliable indication of the effect of helmets. An experimental design, for example a randomised control study, in which the researcher randomly selects who is and who is not going to wear a helmet, is inappropriate for ethical reasons. That is why case-control studies are standard in this area of research .
For biomechanical research, bicycle helmets are tested for shock absorption in a laboratory. During the test, a dummy head with or without a helmet falls down. It is estimated that a bicycle helmet will reduce the risk of serious brain damage in a fall from a height of 1.5 metres from almost 100% to 10%; and in a fall from a height of 2 metres to approximately 30% . In these tests, helmets were used meeting the legal US requirements. The legal requirements for testing bicycle helmets differ by continent (see ) for an overview of helmet standards).
In studies using computer simulations, both the impact on the head as well as the possibly protective effect of a helmet are simulated in a model. On the basis of simulations of three types of single bicycle crashes, researchers conclude that helmet use may reduce the risk of concussion by more than 50% and the risk of a skull fracture by more than 90% . Computer simulation of bicycle-car crashes also show that bicycle helmets may mitigate the severity of brain injuries . Simulation research shows that not only the impact speed but also the impact location on the head determines the protective effect of a bicycle helmet  .
Time series analyses
By means of a(n) (interrupted) time series analysis, the effectiveness of an intervention (for example mandatory helmet use) may be determined by the number of casualties before and after the intervention. Several studies have used this method to determine the effectiveness of bicycle helmets, or rather of mandatory helmet use   . These studies looked at the share of cycling casualties with head or brain injuries before the intervention and during a series of fixed times after the intervention. In general, this kind of study finds a lower effectiveness than studies with a more experimental design. A disadvantage of this kind of study is that they have a prolonged time scale (several years) and that in this period other factors (such as other road safety measures, bicycle use) may also have had an effect on the prevalence of head and brain injuries. Since not every hospital registers whether the injured cyclist wore a helmet or not, this further complicates the design of these studies.