Should I use
blinking bicycle lights?

Betteridge's law of headlines says "No"

Let me start by saying that there is no perfect bicycle light, every situation is different and people like different things. However do make sure that your light complies with your country's laws. Unfortunately many bike lights sold in Europe are not legal to use on a bicycle. Many lack the required certification. Some cheap lights don't even emit enough light.

Many manufacturers include blinking modes because some people think it is safer because people in cars will see them better; an assumption which has not been proven. It is even more far fetched to to state that flashing lights are safer. It's an interesting topic, since it all depends on the conditions and expectations of those involved.

Flashing bicycle lights are legal in some European countries like Belgium, but not in The Netherlands and Germany. Although these countries are a minority they are also the ones that have many cyclists and the most comprehensive legislation when it comes to bicycle lights. What is the reason why they do not allow flashing lights?

Scientific research

Various studies into the differences between flashing and continuous lights in various applications (for example for snow ploughs) have yielded the following conclusions:

  • Flashing lights are easier to notice.
  • It is harder to judge distance of a blinking light.
  • Intoxicated drivers somehow manage to run into blinking lights thanks to the the "moth effect"

In 2008 the Netherlands Organization for Applied Scientific Research (TNO) published a report in which the conspicuousness (noticeability) of various bicycle lights has been tested. As far as research into blinking versus static lights this is as good as it gets.

Please note that that the exact correlation between conspicuousness and safety remains unclear. The report states that there is no research anywhere that shows a direct correlation between the use of bicycle light and safety in traffic (page 14).

Starting at page 7 you can find the summary and conclusions in English. I have quoted it below and highlighted the relevant sections.

Note that the author uses the word "flicking" when refering to a flashing/blinking light.

The first highlighted bullet regarding the front light can be a bit confusing, but what it says is that some static modern lights are more conspicuous than old fashioned ones, while some are less conspicuous. It just depends on the kind of light you have, not on whether it is flashing or not.

Summary

We compared the visual conspicuity of several different types of modern bicycle lights to the conspicuity of some conventional lights, for several relevant traffic scenarios, for different ambient lighting conditions, and for different ways of mounting the lights. We only investigated the detectability of the lighting, because this is a crucial traffic safety issue. Factors like identifiability and distance estimation were not studied.

For front lights we find:

  • Without flicker some of the tested modern front lights are significantly less conspicuous, and others are significantly more conspicuous, than the conventional front light that served as a reference.
  • The conspicuity of a front light that shines straight ahead is independent of the way in which it is mounted (i.e. whether it is attached to the bicycle or attached to clothing).
  • The conspicuity of a modern front light that is not mounted straight (i.e. diagonally, vertically, or slanted) is usually less than the conspicuity of one that is mounted horizontally.
  • For a straight posture of the cyclist a lamp attached to the chest is visible in the entire forward sector.
  • A bent-over/slanting posture can decrease the sector in which the lamp can be seen with more than three times. A bent-over posture also causes the lamp to slant. Since most lamps radiate strongest (have highest luminance) in the forward direction, a slant will also cause the lamp to appear less bright, and therefore less conspicuous.
  • In most cases flicker raises the conspicuity of front lights.
  • There is no clear relation between flicker frequency and conspicuity.

For rear lights we find:

  • Without flicker all tested modern rear lights are just as, or even more, conspicuous than the conventional rear lights that served as a reference.
  • The conspicuity of a rear light that shines straight ahead is independent of the way in which it is mounted (i.e. whether it is attached to the bicycle or attached to clothing).
  • The conspicuity of a modern rear light that is not mounted straight (i.e. diagonally, vertically, or slanted) is usually less than the conspicuity of one that is mounted horizontally.
  • Flicker does not contribute significantly to the conspicuity of rear lights.

The second conclusion seems to contradict the third. What is meant here that a properly mounted light on the bictcle is just as conspicous as a properly mounted light on clothing on the upper body.

Thw third conclusion is about aby improper mounting can result in decreased conspicuity. This depends on the beam pattern of the light. A light that is angled, titled the wrong way can be less conspicous.

Other conclusions:

  • Norms can be defined for the conspicuity of bicycle lights in a given environment, and the corresponding functional requirements for their luminous intensity can be formulated.
  • There is no need to increase the conspicuity of front light by using flicker.
  • Visual perception suggests the following disadvantages of flickering bicycle lights:
    • Flickering bicycle lights make it harder for other road users to estimate the speed and trajectory of a bicycle.
    • Flickering bicycle lights may reduce the conspicuity of emergency services.

Recommendations:

  • Front bicycle lights should be white.
  • Bicycle lights should not flicker.
  • When attached to clothing, bicycle lights should be visible at all times.
  • The importance of proper ways of attachment and alignment should be communicated to the user

Other arguments

A blinking LED uses less energy than a continuously lit LED of the same brightness. Thus bliking lights conserve battery power. This is an undeniable fact, but has nothing to do with safety. If you are worried about conserving battery power there are many solutions, easily rechargeable batteries or dynamo powered lights are the first things that come to mind.

Conclusion

There is too little information to say anything meaningful about how a bicycle light affects the safety of the rider. Those who claim that a particular light is safer than another light often do not have any research to back it up. The reasoning that brighter light equals higher conspicuity and that higher conspicuity increases safety for the cyclist in any meaningful way seems logical, but is not supported by any research.

The fact that it is harder to estimate the distance is a pretty serious argument against flashing lights. The noticeability also depends on the lack of other flashing lights.

Would the road be any safer if everybody would have blinking lights? Absolutely not, it would just turn into a big light show. Of course this argument only bears weight when there are other cyclists around. However, you should not go the road with the assumption that you are the only one there. Attention is a limited resource and by drawing attention to yourself, you are taking it away from others.

Although safety is hard to measure, annoyance is not. There are many reports of people being annoyed by flashing bicycle lights, some claiming even to be temporarily blinded or getting seizures from them.

My conclusion is as follows: there is no evidence that flashing lights are safer, but there is some evidence that it makes people dislike you. My recommendation is to use continous lights, unless you want to annoy other people.