CTL vs ATL Explained for Endurance Athletes (And When the Ratio Lies)

If you have ever opened your fitness-fatigue chart after a hard weekend and felt your stomach drop because the line turned red, I get it. I have had the same reaction. The chart looks precise. The body rarely is.

The short version is the same one I keep coming back to in my own training: CTL and ATL are useful as context, but dangerous when treated as verdicts.

What CTL and ATL actually measure

CTL, or chronic training load, is usually a 42-day exponentially weighted average of daily training stress. It is meant to represent your longer-term training background. ATL, or acute training load, is usually a 7-day exponentially weighted average, so it reacts much faster to a hard block, race, or training camp.

That means CTL is not your literal fitness, and ATL is not your literal fatigue. They are model outputs based on whatever load data your platform receives and however that platform chooses to smooth it.

That distinction matters. If your threshold is wrong, your power data is missing, or your watch switches from one load method to another, the chart changes even if your body has not.

It also matters because the model rewards consistency more than quality. Two athletes can show the same CTL and get there in very different ways. One may be progressing well. The other may just be stacking a lot of moderate work without enough recovery, specificity, or actual performance upside.

I think of the chart as a compressed summary of recent load: CTL shows what you have been sustaining, ATL shows what you have taken on lately, and the gap between them shows short-term strain relative to your own recent baseline.

Why athletes misuse the ratio

From both athlete forums and my own training logs, the same four mistakes show up again and again:

1. They confuse the metric with the thing itself. A low form score or ugly CTL ATL ratio training screenshot is not the same thing as overtraining.
2. They assume the ratio has a magic cutoff. The broader workload literature is much less certain than social media makes it sound, and the chosen time windows are partly arbitrary.
3. They ignore what creates the score. Wrong thresholds, bad TSS, missing power, or a sudden switch from pace load to HR load can distort the chart quickly.
4. They compare raw numbers across athletes or phases. A CTL of 50 can be excellent for one runner and irrelevant for another, and a CTL drop can still coincide with better performance if training becomes more specific.

This is why the ratio lies. It looks clean and quantitative, but it only sees what made it into the model. It does not see poor sleep, underfueling, life stress, illness, travel, or whether your supposedly hard sessions are actually landing.

When a bad ratio does not mean you are overtraining

A bad ratio does not automatically mean you are overtraining if the context says the load was intentional and temporary.

That includes a normal overload week, a training camp, a race block, a big long-run weekend, or the first few days after a hard event. In those cases, ATL jumps faster than CTL by design.

It also does not mean overtraining if the chart is being distorted by settings. This is common in lower-fitness athletes using percentage-based form views, where the same absolute fatigue can look much more dramatic. It is also common after threshold updates or bad data spikes.

And it does not mean overtraining if performance is still stable, recovery markers are acceptable, and you already planned recovery. In many cases that is just functional overreaching, which is a normal part of building fitness.

Real overtraining is bigger than a chart color. It usually involves persistent underperformance plus symptoms that do not clear with a few easy days: poor sleep, mood changes, heavy legs that do not rebound, unusual effort at normal paces, recurrent illness, or clear loss of motivation.

The ratio can also fail in the other direction. You can have a chart that looks fine and still be in trouble because of low energy availability, illness, or life stress. The chart tracks load. It does not diagnose the athlete.

Example using Intervals charts

When I open an Intervals chart that suddenly looks ugly, I check three things before changing the plan.

• Confirm what the chart is actually showing. Form can be displayed as an absolute value or as a percentage of fitness. If your CTL is still modest, the percentage view can make you look far deeper in the red than expected.
• Remember that Intervals and TrainingPeaks do not always display form the same way. Small differences like today-vs-yesterday display logic can create large confusion.
• Inspect the underlying load. Threshold updates, an inflated workout, or a switch from HR-based load to pace-based load can make an Intervals fitness-fatigue chart look dramatic overnight.

That kind of chart may simply mean you stacked a hard four to five days on top of a moderate baseline. If you are still hitting paces, sleeping normally, and already planned an easy 48 hours, the chart is describing accumulated load, not proving overtraining.

The more useful question is not “Is the ratio bad?” It is “Does this chart match what my legs, my workouts, and my recovery markers are telling me?”

Simple decision framework

This is the checklist I would actually use before cutting a workout or rewriting a week:

1. Validate the chart. Check thresholds, load source, missing files, and any suspicious one-day spike.
2. Read the block. Ask whether the fatigue was planned: race, camp, overload week, or long weekend.
3. Check the athlete. Look at sleep, soreness, motivation, mood, appetite, and whether normal paces feel unusually hard.
4. Check performance. If key workouts are still landing, the ratio is probably context. If performance is falling for more than several sessions, pay attention.
5. Decide small, not dramatic. Keep the plan, trim the next workout, or take one recovery day. Reassess after that.
6. Escalate if the pattern persists. If fatigue, illness, or underperformance hangs around beyond a short recovery window, stop treating it like a chart problem.

That, to me, is the real use of CTL and ATL. Not to hand you a diagnosis, but to tell you when to zoom in and ask better questions.

Sources and evidence

The strongest evidence behind this article is on how CTL and ATL are calculated, the limits of single-metric monitoring, and the need to combine objective plus subjective data. Hard ratio cutoffs for safe versus unsafe training are much less certain.

• TrainingPeaks: Fitness (CTL)
• TrainingPeaks: Fatigue (ATL)
• TrainingPeaks: Performance Management Chart explainer
• TrainingPeaks: Why PMC data can be wrong
• Editorial critique of acute:chronic workload ratio
• Systematic review on subjective versus objective monitoring
• Intervals.icu: Form as a percentage of fitness
• Intervals.icu: Today-versus-yesterday form display
• CTS: CTL limits and why performance can rise while CTL falls