Deflection is flex. All forks do it. A fork with lots of deflection is flexible, and can feel squirrelly or soft for heavier riders, but may be perfectly matched for lighter riders or riders who want a little more comfort. A fork with very little deflection is stiff. A stiffer fork can be jarring over bumps, but often is more precise in handling.

It's not, really. There are other factors that go into a decision of which aftermarket fork to buy, if any. And many people of all sizes enjoy forks that have a wide range of deflections, so there are exceptions to any rules I might try to come up with based on rider weight, etc. But you may be considering buying a new fork, and knowing just how it stacks up against what you have already ridden, helps you predict how the ride of your bike might change.

If you are looking for more precise handling, buy a fork that is stiffer. If you are looking for comfort, buy one that is more flexible (or one that has other design features that provide comfort in other ways). If you are looking for light weight, you must decide how flexible is too flexible for your taste, or you may be interested in which forks have the best stiffness-to-weight ratios. And aerodynamics are almost always more important than other factors if solo speed is your goal.

[Comment from John Allen: If a fork is weaker longitudinally than the bicycle's frame, it can spare the frame from damage in a curb impact or hard landing.. A stiffer fork tends to be stronger, though not necessarily.]

I performed a non-destructive quasi-static deflection test. In other words, I anchored the steerer tube horizontally 5 cm from the fork's crown seat and hung a 47.5 pound weight on the dropouts. In all cases a Dura-Ace hub was installed with the skewer tight. For the Lateral (side-to-side) deflection numbers, the hub axle was vertical and deflection was measured with a dial indicator in the direction of loading. For Longitudinal (in line with the direction of the bicycle's travel) deflection, the hub axle was horizontal and the fork blades curved toward the ground. Again, I measured deflection in the direction of loading.

Torsional deflection is the result of a twisting load on a fork. An example of a torsional load is the effort you make when trying to remove a stuck handlebar stem: you twist the handlebars while holding the front wheel with your knees. I didn't measure torsional deflection because I believe it has almost nothing to do with how a fork experiences loads during real-life road riding. If riding did load the fork torsionally to any significant degree, we would have to wrestle the handlebars to keep the front wheel steering correctly. Even during cornering, a rider applies virtually no torque on the bars. In fact, many well-balanced riders can corner quite sharply with "no hands." You can't apply or resist significant torque without hands on the bars.

[Comment by John Allen: torsional loading does occur in the case of "speed wobble" or shimmy of the front wheel. I suspect that the torsional flex of a suspension fork contributed to a very troublesome speed wobble I once experienced. However, Rinard did not test any suspension forks. With an unsuspended ("rigid") fork, most torsional flexibility is in the bicycle's frame.]

Now we get to the controversial part. For years, framebuilders' conventional wisdom held that a fork ought to be flexible longitudinally and stiff laterally. People thought that such a fork would have comfort (achieved with forward deflection) and precise handling (due to the lateral stiffness). However, I am presenting a new theory: longitudinal stiffness is more important for precise handling than lateral stiffness, and lateral deflection can measure quite high before a rider will experience handling troubles. There are two reasons I believe this is true.

One is that longitudinal loading must be higher than lateral loading. I say must because I haven't measured a fork during riding to be sure. (This is, after all, just a theory.) Because of gravity, the rider's weigh loads the bike almost purely downward, especially in turns. Turning causes virtually no lateral load because a rider leans the bike so that loads are nearly perfectly in plane with the wheels. If he didn't, the bike would fall over. The rider's weight then exerts a larger forward load on the fork than it does a lateral load (on the order of, maybe, three times greater, as a guess.)

Also contributing to the longitudinal load (and very little to any lateral load) are the impacts absorbed when the rider hits bumps. The magnitude and direction of bump forces are related to the speed and direction of the bike, and whether the brakes are applied. Now, we presume that a rider who is in control of the bike always goes forward (never sideways), so the direction of any bump force is fore and aft, not lateral. That takes care of the direction of the bump force. It must be true that the magnitude of the bump force increases with speed. In other words, the faster you go, the harder you hit bumps. No surprises here.

These large fore and aft loads indicate the relative importance of longitudinal stiffness because as the loads increase, the deflection increases, and that changes the fork rake at any given moment. If the fork rake changes, the steering geometry of the bike changes, and the rider has to adjust to the bike's constantly changing feel and response. The only time a rider does load the fork laterally is while tossing the handlebars side to side. This happens out of the saddle of course, during sprinting or climbing, for example.

The other reason I believe longitudinal stiffness is more important than lateral stiffness is that people like the handling of forks that are stiffer in the forward direction, such as steel forks, or the Kestrel EMS fork. Okay, my bias is showing: the Kestrel fork tests well in the longitudinal direction, and I used to work for Kestrel. But seriously, how many people do you know who wish their bike didn't have a Kestrel fork, or who don't like the way it rides? The numbers show that of all the forks I tested, so far only steel forks have matched the forward stiffness of the Kestrel fork.

Some riders seem not to notice anything about their bikes. Other riders notice everything. Some riders even notice differences that don't exist! Truthfully, the differences between forks are pretty small. Nice light wheels make a much more noticeable difference. Padded handlebar tape or slightly more or less air pressure make about as noticeable a difference as a new fork, in my experience. But many riders rave about the improved sprinting a stiff fork gives them, or the new-found confidence on fast downhills. Others love the comfort they can finally enjoy on longer rides after installing a more flexible fork. The bottom line is: you might notice a difference.

Yes and no. There are two sources of discomfort in the road: gross bumps (like potholes or reflector dots you hit at speed) are one. The other is the smaller texture of the pavement itself. This texture produces a high frequency vibration as your tire rolls over the road.

The gross bumps can be made more bearable with a fork that flexes a little more. For extremely bumpy roads (like Paris-Roubaix), there are even suspension forks. But the second source of discomfort, the constant vibration, is harder to address. About the only way a normal road fork can lessen them is by the natural damping properties inherent in the material used to make the fork. Perhaps the best way is to use wider tires, since tires are far more compliant than any rigid fork can be.

Road forks are made of steel, aluminum, titanium, and carbon. Of these, carbon is known to damp vibrations about ten times better than the metals. This damping is the reason carbon forks can be both stiff and comfortable. You will still get the jolt of the big bumps with a stiff carbon fork, but the vibrations will be decreased. And a flexible carbon fork is really plush.

If you ever get the chance, try an experiment to experience the difference in damping: ring a carbon fork (like a tuning fork) by slapping it once into the palm of your hand and hold it near your ear, then compare it to a metal fork. The carbon fork damps out the vibrations within seconds, but the metal fork vibrates noticeably longer. Another good illustration of composite's damping compared to metals is a handrail. A metal handrail, if struck, will ring and vibrate for several seconds, but a wooden one (wood is nature's composite) just goes "thunk".

[Comments by John Allen:

Mechanical resonance in the bicycle frame and fork depend on mass, not only on flexibility and damping. I think that the most significant effect of damping is to reduce the height of a resonance peak, which can magnify road vibration, increasing discomfort.

Damon Rinard provided no explanation for the graph below, which was in one of the two versions of this article on our site -- or maybe Sheldon deleted the explanation. The graph is a plot of forward (longitudinal) flexibility of forks against lateral flexibility. For most forks, they are roughly proportional. (Note that flexibility increases toward the right and top of the graph). The Kestrel fork and the Trek Classic Carbon fork both have an intermediate level of lateral flexibility, but the Kestrel fork has unusually low forward flexibility, the Trek Classic Carbon fork, unusually high. A traditional steel track-bicycle fork with round blades would have an unusually high ratio of forward to lateral flexibility, but apparently Rinard did not measure one.

The scale of the graph is in inches and that of the table in the next section of this article, in millimeters, but both are for the same measurements.]

Brand, Model are the brand and model of the fork tested.

Lateral is the lateral deflection in millimeters.

Longitudinal is the longitudinal deflection in millimeters.

Weight is the weight of the fork in grams, normalized to a standard 175 mm steerer tube length.