The higher you can go, the better. There will be less strain on your legs from pushing a bigger gear than you need to. If you bounce or feel like you're wasting energy, then you're going too fast. The more you train, the more sure you'll be of where your cadence belongs. Everyone's different and your number can change. I train at somewhere around 100 on the bike and 96 on the run, but I race more like 105 on the bike and 100 on the run. Last year I probably biked at 90 and ran at just under 90.

If you're looking for scientific, then the power meter can be used, but your level of accuracy will depend a lot on how rigorous you want to be with the data gathering.

If you ride the same course a couple times, at the same time of day, in the same weather, with the same rest -- and then do it a couple more times at different cadences -- you can plot avg power vs. cadence and find your peak. I suppose a trainer would be a great place for this, but I would want some indication of real terrain as well.

However, you would probably also have to be sure to pedal CONSTANTLY at this cadence such that the cadence plot is a flat line with no 'coasting.' I tapped my last 30 rides into excel and plotted Power as a function of cadence and, gee whiz, the higher my cadence, the higher my average power. Without subtracting out time spent coasting, this is meaningless.

Then I got a little geekier, just to see if any meaning could be extracted from normal ride logs, and loaded the raw data from my most recent ride into excel. I made a sort of histogram, calculating the average power produced within each 'bin' of cadences (i.e. avg power between 70-75rpm, avg power 75-80...). The resulting graph shows that my best power/cadence is at 60-65rpm, which I know is not true IN GENERAL. I'm sure this just reflects the fact that I climbed a 1200ft mountain steep enough that I was grinding along at 65rpm. Loaded in a flatter ride that I did TT style - same uncorrelated result. Plotted average speed per bin rather than power and got a much nicer result, showing a bit more of a "Bell curve." This still doesn't account for the rolling terrain, gearing, etc. though, so I'm not convinced. I think with data from specific test rides it could be a useful method of analysis, but getting the data is still the hard part. Perhaps if I concatenated the data from ALL of my rides, I might get something a bit nicer, but that'll have to wait - loading and crunching those numbers would take me a good long while.

I'm sure you could probably model this with a super complex equation for body dimensions and such, but as I think about it, you would really need to even know things like how heavy each segment of your leg is (to calculate things like the force required to overcome the inertia of your thigh at the top and bottom of each stroke!) so it's probably pretty unrealistic to come up with one.

In the end, everything I've ever seen seems to lead back to the "self selected" cadence as being the best for the individual, simple as that. In my experience, it's pretty difficult to try to control your cadence, so I focus more on other features like an efficient "spinning" stroke and keeping aware of when it is or isn't appropriate to coast (i.e. don't coast until you've actually crested a hill and are back up to speed, start soft pedaling before starting up a subsequent hill, things like that)

I hate these type of questions because you just caused me to spend something like the last hour working on this when I have a paper to write :P

Grant

Thanks for the info guys and GGehrke looks like i have some graphs to plot!!!!

But proper spinning technique, especially for new to intermediate cyclist can have a huge impact on cadence. I, for one, believe cadence to be very trainable for newer riders. Spins ups and singel leg drills can take a rider that just didnt know any better pedaling along at 65 rpms, and improve him to a 90 rpm smooth spinner. I know my own cadence when I started riding was around 80, now my "self selected" cadence is 90- 100.

I've been thinking of this alot and trying to record data that will be helpful to this cause. The thinking seems right to me here, that as triathletes, we'll be better served with higher cadences to use more slow twitch on the bike to save the legs for the run.

That being said, we all probably have a measure of our optimal cadence at the moment - I'm much like GG and RV above - when I first hopped on my bike last summer training for my first TRI I was really happy at about 75 RPM. After much work during this base training season, I'm much smoother and happier at more like 85 RPM. I haven't, until now, measured my output at the cadence to determine if I'm doing better or worse at the higher cadence. It seems it would be helpful to have as much power as possible at a higher cadence to maximize the benefits.

So here's my data gathering plan, which I'd like to start as a community data gathering exercise if those with power meters are willing:

1 - only gather data from rides on my trainer for consistency
2 - Select a gear that I can spin a) @ 80 RPM with some comfort, but still a little work and b) up to 105 RPM without bouncing around. (off hand, given that I'm not at my bike right now, I think this would be 2nd biggest ring on cassette and on the large ring up front)
3 - do a 5 min warm up
4 - Intervals:
a.) 2 Min Spinning at 78 - 82 RPM
b.) 2 Min Spinning at 83- 87 RPM
.....
f.) 2 Min Spinning at 103 - 107 RPM

Once I've gathered say 10 or 12 of those rides, then I'd think some meaningful averages could be run of the power at each cadence range.

I invite comments to the data structure first, any thoughts or suggestions are welcome, the hope that a few others might gather the same data to see 1.) if its helpful 2.) see if the combined data yields any particular insights.

If nothing else, would be a 17 min easy ride that might help determine a sense of change in ability to generate power.

Thoughts?

I'm more than a little late to this party, but thought that I would throw out an interesting link... Take a look at this thread at cycling news:
http://www.cyclingnews.com/fitness/?id=2004/letters09-20

In the above discussion is an interesting little fact/opinion: "The hour record which involves no forward acceleration after the start has been many times in the past 100 years, and always with a cadence between 103 and 108 rpm. That suggests to me that this is the optimal cadence for sustained, very steady efforts."

Though, in the following paragraph the author goes on to say: "Ironman Triathlon winners use much lower cadences than road cyclists, typically around 80 rpm. In Ironman Triathlon, efficient use of fuel and oxygen is a key element of success. This type of efficiency is optimized at cadences lower than those typically employed by road racers, and Ironman cyclists, even those who win, don't ride quite as fast as equally elite athletes who compete exclusively on the bike."

I feel the most comfortable at somewhere between 105-110 rpm, with a slight drop on climbs. I don't have a powermeter, but with several years of cycling and many thousands of miles I feel pretty comfortable saying that this is where I am at my best.

Dave

I recall Dave Scott saying on one of his triathlon training tips videos he did for active.com....most of them are on Youtube (and very helpful).....86-92rpms on the bike was ideal for most triathletes. I have also heard that your biking cadence should mirror your running cadence, as Gator said above.