Chronic Training Load (CTL)

Chronic Training Load (CTL) is a slow-moving exponential average of daily training load. It summarizes the longer load history entering the model; it does not directly measure fitness, adaptation, health, or race performance.

The UI may label CTL as Fitness because that is common PMC terminology. Read the value as modeled longer-term load context.

Example values are for UI preview only.

What enters CTL

Trainingload.ai first resolves a daily load value from the available activity load sources and your configured priority. Depending on the activity and settings, that evidence can come from power, heart rate, pace, or Session RPE.

Consequences:

• Missing or changed source data can change the series.
• Editing thresholds or load-source priority can change derived activity load and later PMC values.
• Mixing inconsistent sources can make two periods less comparable.
• A day without recorded load contributes zero to the daily input and the prior CTL decays.

See Training Load before comparing CTL from different setups.

Formula used by Trainingload.ai

Trainingload.ai uses a standard exponential moving-average smoothing factor with a default CTL period of 42 days:

λ_CTL = 2 / (42 + 1)
CTL_today = daily_load × λ_CTL + CTL_yesterday × (1 − λ_CTL)

The 42-day setting controls smoothing; it is not a hard window that suddenly discards day 43. Recent load has more influence and older load decays gradually.

How to read the trend

• Rising CTL means recent daily load is generally above the current CTL level.
• Stable CTL means recent load is roughly maintaining the modeled level.
• Falling CTL means recent load is generally below it, including during reduced training or tapering.

Compare the slope with the activities and plan period that produced it. Two equal CTL values can come from different sports, intensity distributions, load sources, and session structures.

Useful questions

CTL can help answer:

• Has modeled load been building, holding, or declining across several weeks?
• Did missed sessions or a recovery period change the longer trend?
• Is the current plan producing the intended load direction?
• Did a threshold or source change alter comparability?

CTL cannot answer by itself:

• Did the athlete adapt successfully?
• Is tomorrow's hard workout safe?
• Is the athlete recovered, healthy, or race-ready?
• Is one athlete fitter than another?

Use performance trends, completed-workout quality, subjective state, and ATL and TSB context before making a decision.

Avoid universal targets

There is no universal “good CTL,” safe weekly increase, or ideal amateur range. Absolute values depend on the load scale, sport, source, history length, and individual training context. A fixed target copied from another athlete or platform is not a reliable prescription.

Use your own consistently calculated history. When the slope changes sharply, inspect the source activities and the plan rather than diagnosing risk from CTL alone.

Verify a CTL interpretation

• The sport and date range match the question.
• Activity load sources and thresholds are consistent across the comparison.
• The underlying activities are Ready and free of obvious duplicates or sensor errors.
• The conclusion describes load direction, not unmeasured adaptation or readiness.
• A plan decision also considers recent load, execution, and subjective context.

Related pages

• Acute Training Load (ATL)
• Training Stress Balance (TSB)
• Training Analysis Workflow
• PMC Marathon Adjustment