U.S. Roundabouts can Create Risks and Confusion

On September 18, 2018, Maine DOT published the following wording and roundabout design on their Facebook page, but unfortunately the layout is unsatisfactory:

“It’s National Roundabout Week! Roundabouts have proven to be far safer than traditional intersections, but some people are still unsure of how to navigate them…” 

The problem is that, like most other states, Maine is apparently following the Federal Highways Administration [FHWA] ethos on roundabout construction but such guidelines deliberately ignore global best practices that have been developed over the fifty years in which the USA failed to build what are properly called “modern roundabouts.”  Sadly, the result is roundabouts that can have multiple potential safety flaws, just like the one in this illustration, as posted by Maine DOT.

The full layout of a new roundabout, as published by Maine DOT on September 18, 2018.

The problem is that roundabouts need to be versatile to allow for different sizes and the number of exits.  As a result, drivers very much need to be taught a set of simple but universal guidelines that will keep them both safe and confident no matter how large or apparently complex any roundabout might be.  [Naturally, we teach these on all applicable ADoNA defensive and advanced driver safety courses, throughout North America.]  Yet it is the necessary education aspect that the FHWA has decided to omit, and the resultant exclusion from state drivers’ manuals of the best techniques will continue to perpetuate problems throughout the USA.

As with all so-called “turbo roundabouts,” the existing plan is that drivers are guided to enter in the correct lane for the exit they require but typically this only works if a driver already knows which exit they need because advanced layout signs are usually notable by their absence — a classic case of false economy.   If the correct-lane-on-entry stage is achieved, any arrows painted on the road surface should then guide each driver to the required exit for their eventual destination. However, if for example a roundabout has more than four exits, something that roundabouts are particularly good for accomplishing, the lanes-and-arrows method can quickly become unworkable or unsafe, and that is the case in the example in this article.

Let’s look at the above diagram, in numbered sections, to see where problems can arise.  All roundabouts are most easily described as though they were a clock face, so from whatever direction you are approaching it, the “bottom” of a layout diagram is always six o’clock, and the top is always twelve o’clock.

On the above diagram, we are going to start at the five o’clock position (bottom right) and work our way around in a counter-clockwise direction, obviously the same way as traffic circulates.

1.  Roughly the “five o’clock” section of a roundabout diagram published by Maine DOT on September 18, 2018.

What we can see from the white arrows on the entrance to the roundabout, at the bottom of the image, is that drivers can use either the left-hand or center lane to enter the roundabout if they wish to go beyond the first available exit,  However, it is what can happen after they pass that first exit that matters, because of the two lanes heading farther around.  (For those who are into technicalities, there is another significant flaw shown in this section.  To read what that is, please scroll down to the ‘Footnote’.)

2.  Roughly the “one o’clock” section of a roundabout diagram published by Maine DOT.

Take notice of the curved arrow closest to the bottom edge of this section of the diagram.  In accordance with the earlier, curved arrow [on the illustration of section 1], a driver might understandably have selected this lane if s/he wanted to use the same exit as the orange car and orange line (because at that point the orange car’s destination was not the first available exit. So NOW WHAT?  Admittedly, a bit of cautious driving would probably save the day but this situation, which should never arise, has happened because:

  1. It is sometimes unwise and even unsafe to try to “turbo” a roundabout with more than four exits;
  2. Drivers in the USA are having ALL proper, best-practise roundabout advice withheld from them because somebody in Washington DC doesn’t know what s/he is doing and thinks they can re-invent a highly-proven method … but s/he failed!

Now see how both arrows near the left-hand side of the above image point for a straight-ahead exit (correctly) but the one in the left lane also points left for drivers who wish to keep on going around to later exits.

3.  Roughly the “ten o’clock section of a roundabout diagram published by Maine DOT.

Here are those same arrows again, but now at the top right-hand corner of this illustration, and we can see that the driver of the red car is in the correct lane if s/he wants to take the next exit.  But if the driver does not want the next exit and needs a later one, and if they are not very familiar with roundabouts, or are nervous, or don’t exactly know which exit they need until they reach it (which is a sadly common situation at badly signed roundabouts), which lane does s/he choose?  The left one or the right?   Another potentially dangerous situation is developing.

4.  Roughly the “eight o’clock section of a roundabout diagram published by Maine DOT.

So here, let’s imagine that despite the red line. the driver of the red car is not going to use the “six o’clock” exit at the bottom of the picture but is going farther around the roundabout.  Look at the road arrows.  It is perfectly legitimate for the red car to be in either lane.

5.  The “six o’clock section of a roundabout diagram published by Maine DOT.

Now imagine that the driver of the red car is in the left-hand lane of the roundabout and is in close proximity to the blue vehicle (or any other vehicle that is in the right-hand lane), but the red car is heading for the exit that goes out of the right-hand side of this part of the image while the blue car is following its blue line….. Once again, NOW WHAT?   The same highly undesirable situation has arisen as described under the image number 2, above.

The good news is that some American states are doing a better job of it, and an important part of this is to have a large, clear ‘map’ sign well before each entrance to any roundabout, showing the destinations of all the exits.  This, in itself, makes lane selection utterly simple, as long as drivers have been taught correctly how to drive around ANY roundabout, and that education needs to include:

  • Early and safe selection of the correct lane, before even reaching the roundabout, based upon an immensely simple rule;
  • Signalling one’s intentions and direction, as applicable, before entering the roundabout;
  • Signalling as appropriate while going around the roundabout;
  • Giving a right-turn signal at the appropriate moment — not too early and not too late –to exit the roundabout;
  • Safely carrying out a gradual lane change, if necessary, when heading for the correct exit.
A six-exit roundabout in Washington State, where rather than trying to herd drivers to follow arrows (rather than look where they are going), the correct choice of lanes can be made simply by using proven guidelines and seeing the map on this sign.
An excellent ‘map’ sign, showing the layout and exit-destinations of a five-exit roundabout in Washington State. (This is clearly a roundabout where painting arrows on the road/pavement surface would be confusing and useless.)   Copyright image, 2015.

If that list of bullet-points sounds complicated or excessive to you right now, don’t worry.  It does to anyone before they learn it but it turns out to be utter simplicity.

On the whole, however, this very necessary education simply has not happened here in the USA and the regrettable outcome is that, having eventually decided to utilize modern roundabouts, which were first invented in Britain in the 1960s, the FHWA has inexplicably chosen to make things more confusing and more difficult for American drivers.

If, through anxiety, drivers simply stay in the right-hand lane all the way around, as many already do, then delays, anger and even collisions will result, and an excellent resource is being wasted.

To paraphrase Maine DOT’s words, the FHWA’s flawed decision makes it anything but surprising that many American drivers will remain “unsure of how to navigate [roundabouts],” and remain unnecessarily apprehensive about them, too.

We have several posts on different aspects of this topic.  To easily locate them, simply go to the search box at the top, right-hand corner of any page on this website and search for roundabouts.



In the sectional illustration number one, the “straight ahead arrows under the orange line and the blue line cannot possibly be correct.  (See the full image illustration above if you don’t understand why.)  The arrow in the middle lane, under the orange line, should be a three-way arrow showing left, straight and right.  The arrow in the right-hand lane, under the blue line, should be a right-turn-only arrow because it is leading to a compulsory exit.


With the specific exception of the diagram published by Maine DOT, and smaller sections of that diagram, above, please be aware that this website is registered with the United States Copyright Office and that punitive legal action for damages may be taken against anyone who breaches our copyright. This, however, does not stop you from posting links to any of our pages, and you are welcome to do so.

Author: EddieWren

Eddie Wren is the CEO and Chief Instructor at Advanced Drivers of North America. His driver safety background is given at: http://www.advanceddrivers.com/ceochief-instructors-resumecvbio/

6 thoughts on “U.S. Roundabouts can Create Risks and Confusion”

  1. As one who has been involved with the design of around 50 roundabouts and active with the ITE and TRB roundabout groups, I offer the following comments. I also have driven through approximately 1500 roundabouts in 26 countries with both right and left side driving. I have driven through roundabouts in 21 U.S. states.

    With respect to the Maine roundabout; having more than 4 legs at a multi-lane roundabout arguably makes the design, striping and signing very tricky. Unfortunately, the use of roundabouts at locations with more than 5 legs is often the suggested solution preferred by landscape architects, planners and politicians. Every effort should be made to remove one of the legs and have it connect to one of the remaining legs at a secondary “T” intersection. However, that cannot always be done.

    The idea stated in your associated text with the Maine information is that drivers should be allowed to circulate in the outside lane until they reach their exit; this creates several problems. That concept really only works for single lane roundabouts. For roundabouts needing more capacity, multi-lane entrances and exits will be needed. If outside lane traffic is not only allowed, but encouraged to pass through more than one two lane exit, there is a greatly increased crash potential for that improperly circulating traffic to side swipe a vehicle properly using the left lane of a two lane exit. If drivers do this, according to most state traffic laws, they are changing lanes by crossing the lane lines and therefore have to be aware to traffic in the adjacent left lane and yield to them. The proper way to mentally envision a correctly designed four-leg, two-lane modern roundabout is that it is a very compact junction of a set of 4 one-way to one-way intersections with an island in the center. IT IS NOT A CIRCULAR ROADWAY CONNECTING A SERIES OF “T” INTERSECTIONS! That is an old style rotary and if it has two lanes and all exiting is expected from the outside lane, what’s the left lane to be used for then; passing? Again, having more than 4 legs does make the design more challenging, but not impossible.

    Roundabouts in the U.K., after they went to Yield on entry in the 1960’s, became popular. Until recently, they didn’t stripe lanes on their approaches or in the central roadway area. They had empirically developed formulas that gave them their needed approach widths based on traffic forecasts and the central roadway width was set at 1 to 1.2 times the width of the widest approach. The roundabouts in Vail, CO were originally built and operated that way. The joke was, U.K. roundabouts operated like city streets in Italy. The number of lanes was determined by the width of the vehicles using it at that time. The diagrammatic sign shown in your comments, was all that was felt to be needed. They used the “door handle” rule. If you cloud see the door handles of the vehicle in the adjacent lane, they had the right-of-way over your vehicle.

    The U.S. and Canada went on a different path. However, in recent years in the U.K., first striped lanes on the approaches were implemented and then eventually striping in the central roadway. Like North America, they now do expect drivers to get in the proper lane and then follow that lane until their exit. Sometimes they use a spiral striping plan, where lanes are added on the inside and then drop on the outside.

    One of the most frequent roundabout design shortcomings (not a flaw sufficient to generate lawsuits, however because there may extenuating design constraints) is the failure to make the splitter islands wide enough in order to reduce the distance between a roundabout entrance and the next exit. Having a shorter distance at this point increases the entrance angle to improve the driver’s view of approaching traffic and it also reduces the distance between entrances and exits to reduce any temptation to weave or change lanes. Having wide splitter islands allows entering traffic to know sooner when an approaching vehicle is using the previous exit. The Maine roundabout design does a fairly good job of minimizing the conflict areas.

    Unfortunately this comment box does not allow the addition of pictures. I would be glad to send you photographs and drawings of the proper way to design and stripe a modern roundabout.

    By the way, the Maine roundabout is not a true “Turbo Roundabout” at all. It is a partially spiral striped roundabout. The turbo roundabout is a specific roundabout design technique developed in the Netherlands by Lambertus Fortuijn. It features perpendicular approaches and exits connecting to a large circular junction with raised islands in the central roadway to require traffic to stay in their lanes to complete their left, through, and right movements. Often part time signal control is added to turbo roundabouts.

    The turbo design was developed to overcome the historic problems associated with the perpendicular approaches and exits found at most mainline European “roundabouts” which are really traffic circles and are referred to as “concentric designs”. The typical European design methods encounter crash problems when there are multiple lanes. The German design practice is that they may occasionally have more than one lane in the roundabout, but only single lane exits are allowed. In Germany, there can be no three lane roundabouts. These policies limit the use of roundabouts to lower volume intersections and forego the reduction in severe crashes that a roundabout can provide when compared to a traffic signal. That is what led the Dutch to develop the “Turbo”.

    If one wants to rely on the years of experience in roundabout design from outside of the U.S., one should look to the roundabouts in Australia and New Zealand.

    I once saw graffiti painted on a building when getting off the ferry across the English Channel to France. It read, “Hell is a French roundabout”. The French roundabout designs and their drivers have gotten better since then.

    So for the roundabout from Maine shown in the attachment, there definitely are striping issues. The lane designations in the one o’clock approach shows 2 lanes with both left and through moves allowed. The left lane on that approach should be designated with a hard left, a soft left and a through arrow, the center lane as a through only.

    If the 11 o’clock approach and 4 o’clock exit are the same named roadway then the lane designation should be hard left only in the left lane, hard left, through and soft right in the center lane and right only in the right lane. The lane designations in the 6 o’clock approach are also ambiguous. The lane use arrows should show hard left only in the left lane, hard left, soft left and soft right in the center lane and hard right and soft right in the right lane. In the roundabout at the 4 o’clock position, the pavement arrows are correct with left in the left lane, left plus through in the center lane and through in the right lane. The lane use arrows on the 8 o’clock approach would be dependent on if the 8 o’clock to 1 o’clock roadway is a named through route. If so, the pavement arrows should be left and through. The right lane should be through and soft right. In the 7 o’clock segment of the roundabout itself, the pavement arrows should be through in the left lane and through plus soft right in the right lane. Pavement arrows within the roundabout should only show which moves are allowed at the next diverge point. Pavement arrows in the approach lanes should show all allowed moves. Diagrammatic and lane use signs on the approaches should match the approach arrows.

    Many agencies in the U.S. do not use the center island dots and some don’t use fish hook arrows.

    If you would like to see examples of properly designed and striped multi-lane roundabouts, please send me a link to post those pictures to.

    This is a very timely subject, but the FHWA and roundabout designers in the U.S. are not going back to building rotary junctions with vehicles running laps in the outside lane.

    1. Thank you for your extremely detailed response, Dennis. It will take me some time to respond in anything like adequate detail!

      One point I will make now, however, is that — as mentioned in another response — after a long day yesterday, I did accidentally hit “post” for my article rather than “save draft,” so it was published too early, before satisfactory completion. You will now find that there have been several modifications to the text.

      If you understood that “The idea stated in [my] associated text with the Maine information is that drivers should be allowed to circulate in the outside lane until they reach their exit,” my wording must have been accidentally very misleading (and I will take another look at it). I am hugely in favour of multi-lane roundabouts, so everyone using the outer lane would be anathema to me.

      P.S. Incidentally, the best person with whom you might like to discuss the finer points and concepts of roundabout design in this discussion is Lance Fogg, a very experienced highway engineer from Britain.

  2. It’s actually quite an old design with the entries coming in with a low angle of deflection, almost tangential to the circulatory carriageway. The key issue for safety at roundabouts is “slow in, fast out”. There needs to be smaller radii and greater deflection on entry . Drivers cannot and will not twist their heads to properly assess oncoming traffic from the right. The approach needs to be more at 90 degrees to the tangent so they can better judge traffic from the right.
    The example seems to have made use of spiral markings. In this case the markings are not consistent across the junction so I have to assume they have been tailored to the predominant flows. I’ll give it 6 out of 10 for trying.

    1. Thanks for your input, Lance, but do you mean from the left, rather than “so they can better judge traffic from the right”?

      P.S. Lance, may I ask you to take a look at the replies I have received from Scott Batson and Dennis Eyler, please?

  3. People get confused driving??????? Do tell.
    Your definition of potentially dangerous is not defined. Crashes at 15-20 mph are rarely dangerous to occupants of automobiles. Hence the significant safety record of modern roundabouts in the US.
    I do hope that, when ‘educating’ your paying customers about modern roundabouts, you point out the safe practices (and often laws) regarding yielding to all circulating lanes and never passing in the circulatory roadway.
    I don’t know many people that drive without knowing where they are going, unless they are visitors, or new to an area, so your premise is basically flawed.

    1. Mr. Bateson, I think it fair to say that “potentially dangerous” means a situation involving risk of a collision, but it also includes any scenario which could result in a driver having to brake hard or swerve.

      I’m sorry that you feel the need for sarcasm about whether we teach about yielding when entering roundabouts. The answer is that yes, of course we do, but we teach a two-way rule that also minimizes delays and increases traffic flow, which does seem to be the sometimes rather obsessive goal that traffic engineers hold. If you wish to see my credentials for designing our training curriculum, I invite you to see my résumé at: http://www.advanceddrivers.com/ceochief-instructors-resumecvbio/

      As for “the safety record of roundabouts in the US,” I need to point out that I was driving on modern roundabouts in Britain over thirty years before the USA even got around to building them and I was enforcing the laws of the said modern roundabouts from the mid 1970s onwards, as a traffic patrol police officer (and investigation crashes, too), so yes, I do also understand what safety and good design are in relation to this topic.

      And, of course, yes visitors and new residents in any location have no clue about the layout of intersections up ahead, so unlike you we DO care about such people by supporting removal of the flaw that presumptuous engineers inject when they don’t give sufficient warning of which road goes where. Assuming your ‘PE’ is still active and you are not retired, that’s collectively your “bad,” not ours.

      On a closing note, however, the article has been updated since you saw what was effectively an unfinished version, published a little too early due to my own error, after a long day, of clicking ‘publish’ rather than ‘save draft.’ I hope it is now clearer as a result of the update.

      P.S. Incidentally, a good person with whom you might like to discuss the finer points and concepts of roundabout design vis-à-vis Britain/USA, within this discussion, is Lance Fogg, a very experienced highway engineer from Britain.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.