The definitive guide on how to best use your power meter, both in training and racing. Co-authored by Hunter Allen, founder of CyclingPeaks Software and noted physiologist Andrew Coggan, Ph.D. This book is the perfect companion to WKO+. Buy Now direct from us to get a signed copy!

Watts per Kilogram: Using the CompuTrainer Indoor Ergometer to Improve Your Performance by Richard Wharton. This e-book contains over 125 pages dedicated to understanding just how powerful you can become with a CompuTrainer as part of your program. The ultimate companion to ERG+, CRS+ and Real3D.

Threshold power: what is it, why is it important, and how do I measure it?
Andrew R. Coggan, Ph.D.

For more than 30 years, exercise physiologists have known that the exercise intensity at which lactate begins to accumulate in a person's blood - that is, their lactate threshold (LT) - is a powerful predictor of their endurance performance ability. This is because although an individual's cardiovascular fitness, i.e., their maximal oxygen uptake (VO2max) sets the upper limit to their rate of aerobic energy production, it is their metabolic fitness, i.e., their LT, that determines the percentage or fraction of their VO2max that can they can utilize for any given period of time. The physiological factors determining LT are complex, but in this context blood lactate levels essentially serve as an indirect marker for biochemical events within exercising muscle. More specifically, a person's LT reflects the ability of their muscles to match energy supply to energy demand, which in turn determines the fuel "mix" (i.e., carbohydrate vs. fat) used and the development of muscle fatigue. Consequently, LT - especially when expressed as a power output, which also takes into account cycling efficiency - is the single most important physiological determinant of performance in events ranging from as short as a 3 km pursuit to as long as a 3 week stage race. Just as importantly, because the metabolic strain experienced when exercising at a given intensity is dependent upon the power output relative to power at LT, this parameter provides a physiologically sound basis around which to design any power meter-based training program.

CyclingPeaks Software explicitly recognizes the crucial importance of power at LT by allowing you to enter a value for your current "threshold power" (and threshold heart rate) into your "Athlete Settings" file. This value is then used to calculate the intensity factor and training stress score for every file you analyze [see "What are normalized power, intensity factor (IF), and training stress score (TSS)?"]. In addition, if you wish CyclingPeaks Software will use your threshold power to automatically calculate seven suggested training ranges, or levels. Alternatively, CyclingPeaks Software also allows to custom define your own power-based training levels.
So, how do you go about determining your threshold power? Obviously, one way is via laboratory testing with invasive blood sampling, but few people have access to such testing on a regular basis. In addition, power at LT as determined in this manner is often significantly below what athletes and coaches tend to think of as a "threshold". A more convenient and possibly more accurate way of determining your functional threshold power is therefore to simply rely on data collected using your power meter in the field. There are a number of different ways of doing so, each of which has its advantages and disadvantages, but all of which provide very similar estimates of threshold power. In order of increasing complexity, these are:

1. A good estimate of your functional threshold power can often be obtained by simply uploading all of your training data into CyclingPeaks Software, and then examining the power frequency distribution found on your "Athlete Home Page". Because exercising above threshold power is quite strenuous and there is a limit to how long you can do so, there will often be a rather noticeable drop-off above this point in this graph. (This same approach works even better for identifying an individual's spontaneously-achieved maximal heart rate - thus reducing or even eliminating the need for formal testing!) Of course, this method works best if the time period being examined includes some high intensity training and/or racing, which serves to make the distinction between sub-threshold and supra-threshold efforts more distinct. Also, sometimes the drop-off in time spent above threshold power is more apparent when the width of each power "bin" is reduced from the default of 20 W to a smaller value, e.g., 5 or 10 W. CyclingPeaks Software has been specifically designed to allow you to customize graphs, to make such analyses possible.
2. Another way of estimating your threshold power without performing any formal testing is to simply evaluate the steady power that you can routinely produce in training during longer hard efforts, e.g., intervals or repeats aimed at raising LT, or during longer climbs. In CyclingPeaks Software, perhaps the easiest way of doing this is to add a horizontal grid line to a "stacked" graph of an appropriately-chosen workout (or race), and looking for places where your power is quasi-constant for some minutes at a time. You can then adjust the gridline up or down as needed to hone in on the best estimate of your threshold power.
3. Perhaps an even more precise way of determining your threshold power, yet one which still doesn't require any formal testing, is to examine your normalized power during hard ~1 hour mass start races. Since CyclingPeaks Software automatically calculates normalized power even if you haven't yet entered a value for your threshold power, using the program to first analyze several race files may be the quickest way to deriving a good estimate of your threshold power.
4. Since by definition the best measure of performance is performance itself, the most direct estimate of your sustainable (threshold) power will be obtained by simply doing a ~1 hour TT. By examining the horizontal graph of the data from such a TT in CyclingPeaks Software (perhaps with a little smoothing applied), you will be able to quickly tell whether your effort was well-paced, or if perhaps you started out too hard and then later faded, resulting in the average power somewhat underestimating your true threshold power.
5. Finally, those who are more mathematically inclined may wish to perform formal testing to determine their "critical power" as described in the scientific literature. Briefly, this approach consists of plotting the total work performed (in joules) during a series of relatively short (i.e., between 3 and perhaps 30 min), all-out efforts against their duration (in seconds), then fitting a straight line to the data points. The slope of this line is critical power, which corresponds quite closely with functional threshold power determined using any of the previously-described methods.

Since one goal of any training program is to increase power at threshold, the value you have entered into CyclingPeaks Software should be periodically reassessed to be certain it is still accurate. (In particular, an intensity factor of more than 1.05 - meaning that normalized power is more than 5% greater than threshold power - during a ~1 hour mass start race is often evidence that threshold power is greater than the value entered into the program.) How often threshold power will change significantly will depend in part on an individual's training history and habits - for example, someone who is just beginning in and/or returning to cycling may see large and rapid changes in their threshold power, whereas an experienced rider who has been training for many years and/or an athlete who maintains a high level of conditioning year round will probably experience much less variation. In general, however, assessing threshold power a few times per year (e.g., near the start of training as a baseline, partway through the pre-competition period to track improvement, and during the season to determine peak fitness achieved) is probably sufficient.