External Load and Internal Load: Concepts, Differences, and Monitoring Strategies

Prerequisites: This article assumes familiarity with basic energy systems, cardiovascular responses to exercise, and fundamental training principles. If any of these topics are new to you, start with:

• Understanding Energy Systems: ATP-PCr, Glycolysis, and Oxidative Phosphorylation
• Cardiovascular Responses and Adaptations to Exercise
• Training Principles: Overload, Specificity, Reversibility, Individuality, and Progressive Overload

Learning Objectives

• Define and distinguish between external load and internal load as separate constructs.
• Classify the key metrics and tools used to measure external and internal load.
• Explain why internal load drives training adaptation and how integrating both load types enriches interpretation.
• Recognise the scientific limitations of the term “training load” and apply principles for appropriate terminology use.
• Design a practical monitoring strategy that integrates external and internal load in the field.

---

What Is External Load?

External load is the physical work prescribed to or imposed on the athlete, independent of the athlete’s internal response. In a resistance training session, external load refers to the weight lifted, sets, and repetitions. In team sports such as football, it refers to the distances covered, speeds reached, accelerations performed, and other locomotor outputs recorded by tracking systems (Impellizzeri et al., 2019).

The critical feature of external load is that it describes what was done, not how the athlete responded. Two players completing the same small-sided game cover similar total distances and perform comparable accelerations, yet their physiological cost may differ substantially. External load captures the stimulus as it was delivered.

Common external load metrics in football include total distance (TD), high-speed running distance (HSR), sprint distance, acceleration and deceleration counts, and composite measures such as Player Load derived from tri-axial accelerometers (Miguel et al., 2021). A systematic review of 82 studies found that distance covered per speed zone (50 studies), TD (47 studies), and acceleration/deceleration metrics (35 studies) were the most frequently reported external load indicators (Miguel et al., 2021).

External load data, however, cannot reveal the physiological cost of the work performed. A winger and a central midfielder may accumulate identical HSR distances in a session, but the metabolic and cardiovascular strain on each player depends on individual fitness levels, fatigue state, and positional demands. External load tells you what the athlete did, not what it cost them. This limitation is precisely why a second construct is needed.

What Is Internal Load?

Internal load is the psychophysiological response that occurs within the athlete during the exercise bout. It represents the biological cost of the external load — the cardiovascular, metabolic, and perceptual reactions that the body initiates to meet the demands of the prescribed work (Impellizzeri et al., 2019).

Heart rate (HR) during a training session reflects the cardiovascular demand imposed by the activity. Session rating of perceived exertion (sRPE) — calculated by multiplying the athlete’s perceived intensity on a 1–10 scale by the session duration in minutes — quantifies how the athlete experienced the session as a whole (Riboli et al., 2023). Blood lactate concentration during exercise provides a snapshot of the metabolic cost at a given moment. Each of these captures a different dimension of the internal response.

A key distinction separates internal load from post-exercise recovery markers. Internal load occurs during the exercise bout. Indicators collected after the session — such as heart rate variability (HRV), countermovement jump (CMJ) performance, creatine kinase (CK) levels, or self-reported wellbeing scores — are not internal load. They reflect the athlete’s state or the training response that follows, not the in-session physiological reaction itself (Impellizzeri et al., 2019). Conflating these categories leads to conceptual confusion that undermines the monitoring framework.

Category	Definition	Example Metrics
External load	Physical work prescribed to the athlete	TD, HSR distance, sprint distance, Acc/Dec counts
Internal load	Psychophysiological response during exercise	HR, sRPE, blood lactate
Athlete state	Post-exercise readiness and recovery markers	HRV, CMJ, CK, wellbeing questionnaires

This three-part classification — external load, internal load, and athlete state — provides a cleaner conceptual map than the traditional two-category model and prevents the common error of labelling recovery markers as internal load (Rebelo et al., 2026).

Internal Load Drives Adaptation

If external load describes what was prescribed and internal load describes how the body responded, the natural question is: which one determines the training outcome?

The answer, grounded in the dose-response relationship of exercise, is that training adaptation is ultimately driven by internal load. The same external load produces different internal responses across athletes depending on their fitness level, nutritional status, psychological state, sleep quality, and genetic profile. Two athletes performing identical interval sessions may experience vastly different cardiovascular and perceptual demands. The athlete with a higher aerobic capacity completes the session at a lower relative intensity, accumulating less internal strain. The athlete with lower fitness works closer to physiological ceiling, generating a larger adaptive stimulus (Impellizzeri et al., 2019).

This principle has direct monitoring implications. When external load is standardised — for instance, a submaximal running test performed at fixed speed — changes in the internal response over time reveal shifts in fitness status. A decrease in HR and RPE at the same running speed signals improved cardiovascular fitness. An increase in internal load at the same external load may indicate accumulated fatigue, detraining, or illness (Impellizzeri et al., 2019; Pillitteri et al., 2024).

The external-to-internal load ratio (EL/IL ratio) operationalises this principle. An athlete who covers more high-speed distance (higher EL) while reporting a lower sRPE (lower IL) demonstrates improved capacity to tolerate the prescribed demands. Conversely, declining external output alongside rising internal load signals a reduced capacity that warrants investigation (Pillitteri et al., 2024).

Neither measure is sufficient alone. Tracking only external load ignores the individual response and treats all athletes as identical. Tracking only internal load sacrifices the precision of the training prescription. Integrating both constructs creates the interpretive framework needed for informed training decisions.

Combining physiological and perceptual internal load indicators can even help distinguish the type of fatigue an athlete is experiencing. When both HR and RPE increase at a standardised external load, muscle fatigue is likely present. When RPE increases but HR does not, mental fatigue may be the primary driver — a distinction that has practical implications for subsequent session design (Impellizzeri et al., 2019).

The Problem with the Term “Training Load”

The term “training load” is used ubiquitously in sport science, yet it carries significant conceptual problems. In the International System of Units (SI), “load” is a mechanical term that refers to force, measured in Newtons. When practitioners speak of “training load,” they are not referring to force. They are referring to volume, intensity, or some combination of the two — constructs that should be distinguished rather than conflated (Staunton et al., 2022).

The issue extends beyond academic pedantry. “Training load” functions as a meta-construct: a single label applied to fundamentally different things. In one study, “training load” might refer to total distance covered (an external, volume-related metric). In another, it might mean sRPE (an internal indicator integrating perceived intensity and duration). In a third, it could denote the acute:chronic workload ratio — a derived metric combining both volume and intensity across different time windows. Using a single term for all of these obscures what is actually being measured and compared (Staunton et al., 2022).

Proprietary metrics compound the confusion. Player Load, a widely used accelerometer-derived measure, calculates the scaled vector magnitude of instantaneous acceleration changes across three axes. Despite being reported in arbitrary units (not Newtons), it carries the word “load” in its name. The algorithm is proprietary, and independent analyses have reported discrepancies between the published formula and actual software output (Staunton et al., 2022).

A practical alternative is the FITT-VP framework (Frequency, Intensity, Time, Type, Volume, Progression) proposed by the American College of Sports Medicine. Rather than bundling everything under “load,” this framework distinguishes between volume (the product of frequency, intensity, and time) and intensity as separate, clearly defined constructs (Staunton et al., 2022).

No consensus replacement term has been adopted by the field. Until one emerges, practitioners should continue to use “training load” where necessary but always specify the context: which metric, measured by what tool, representing which construct (volume, intensity, or a composite). Precision in language supports precision in thinking.

Technologies and Caveats for Measuring External Load

External load measurement in team sports relies on four main technology categories: satellite-based systems (GPS/GNSS), optical tracking, radio-frequency identification (RFID) and ultra-wideband (UWB) systems, and inertial measurement units (IMU) containing accelerometers, gyroscopes, and magnetometers (Clubb & Murray, 2022).

Technology	Environment	Strengths	Key Limitations
GPS/GNSS	Outdoor training	Portable, individual unit per player	Signal quality varies; lower accuracy at high speeds
Optical tracking	Stadium (fixed cameras)	All players tracked simultaneously; ball tracking	Fixed infrastructure; costly; limited to match venues
RFID/UWB	Indoor and outdoor	High sampling rate; works indoors	Infrastructure cost; signal interference
IMU/Accelerometer	Any	Stride-level data; collision detection	Arbitrary units; proprietary algorithms

Buchheit & Simpson (2017) proposed a three-level classification of tracking metrics based on their validity and reliability. Level 1 metrics — distances covered at various speed zones — are derived from all positioning systems and carry the highest measurement confidence. Level 2 metrics — involving acceleration, deceleration, and change of direction — are available from most systems but have lower reliability, particularly at higher intensity thresholds. Level 3 metrics — stride characteristics, Force Load, and other accelerometer-derived variables — are exclusive to wearable micro-technology and show promise for fitness and fatigue monitoring but require further validation.

A critical paradox emerges from this classification. The metrics that practitioners consider most important for performance and injury monitoring — HSR distance, high-intensity accelerations and decelerations, and metabolic power — are precisely the ones with the lowest validity and reliability. GPS inter-unit variability can reach up to 50%, and accuracy decreases as movement intensity increases (Buchheit & Simpson, 2017).

Speed thresholds used to define intensity zones present another challenge. Absolute thresholds (e.g., HSR defined as >19.8 km/h) apply the same cut-off to every player regardless of individual capacity. A player with a maximal sprint speed of 34 km/h and one with 29 km/h are measured against the same benchmark, which may over- or underestimate the physiological cost for each. Individualised thresholds — set relative to each player’s maximal aerobic speed (MAS), anaerobic speed reserve (ASR), or maximal acceleration/deceleration capacity — provide more physiologically meaningful data but introduce additional testing requirements (Rice et al., 2023; Pimenta et al., 2026).

Best practice calls for assigning each player the same GPS unit across all sessions to minimise inter-unit error, and for using system-specific calibration equations when comparing training data (GPS) with match data (optical tracking) collected by different technologies (Buchheit & Simpson, 2017; Murray & Clubb, 2022).

Tools and Caveats for Measuring Internal Load

Internal load measurement centres on three main approaches: heart rate-based methods, perceptual scales, and wellbeing self-reports.

Heart Rate-Based Methods

HR monitoring has been the primary internal load tool in football for over two decades. Because HR reflects the integrated cardiovascular response to exercise, it serves as a valid proxy for aerobic metabolic demand during sustained activity (Rice et al., 2023).

Zone-based HR analysis divides exercise time into intensity bands relative to maximal HR. A more nuanced approach uses Training Impulse (TRIMP), which integrates session duration with HR data. The original TRIMP formula — duration (min) × (HRex − HRrest) / (HRmax − HRrest) — uses the average exercise HR (HRex), which may under- or overestimate the actual internal demand during intermittent activities such as football drills (Rice et al., 2023).

Modified TRIMP (TRIMP_MOD) addresses this by applying exponential weighting factors to time spent in each HR zone, giving disproportionately greater weight to higher-intensity zones. The zone-specific weighting factors are: zone 1 (65–71% HRmax, ×1.25), zone 2 (72–78%, ×1.71), zone 3 (79–85%, ×2.54), zone 4 (86–92%, ×3.61), and zone 5 (93–100%, ×5.16) (Rice et al., 2023).

Additional TRIMP variants exist, including Banister’s TRIMP (bTRIMP), exponential TRIMP (eTRIMP), and individualised TRIMP (iTRIMP) — each reflecting different mathematical approaches to weighting the HR-intensity relationship (Jamieson, 2022). No single TRIMP variant is universally superior; selection depends on available data, sport context, and practitioner expertise.

HR-based internal load measures carry important limitations. HR responds primarily to aerobic metabolic demand and is therefore a valid internal load indicator for endurance-type activities but a poor one for resistance training, where the cardiovascular response does not proportionally reflect the neuromuscular strain (Impellizzeri et al., 2019). In football, drills that emphasise short, explosive actions may not elevate HR in a manner that accurately reflects the total physiological cost.

Full cardiac-autonomic recovery following high-intensity cardiovascular exercise requires a minimum of 48 hours, a timeline that influences how recovery-day HR data should be interpreted (Jamieson, 2022).

Perceptual Measures: RPE and sRPE

Rate of Perceived Exertion (RPE), assessed approximately 30 minutes after a session on a 1–10 scale, captures the athlete’s global perception of session difficulty. sRPE multiplies this rating by session duration to produce a single internal load value. For example, a 90-minute session rated as RPE 7 yields an sRPE of 630 arbitrary units (Riboli et al., 2023).

sRPE is a validated and reliable tool for reporting aerobic training load. It correlates with average HR, acute HR changes during steady-state exercise, and high-intensity interval responses (Riboli et al., 2023). A survey of over 200 football practitioners worldwide found that RPE was the second most commonly collected metric after locomotor variables, used by 45–50% of respondents (Dello Iacono et al., 2025).

RPE and sRPE should be viewed as complementary to HR-based methods, not as replacements. Each captures a different facet of the internal response. When HR monitors are unavailable, sRPE provides a practical fallback for estimating aerobic training load (Riboli et al., 2023).

Wellbeing Self-Reports

Athlete Self-Reported Measures (ASRM) — daily questionnaires covering items such as muscle soreness, fatigue, sleep quality, stress, and mood — provide information about the athlete’s subjective state. These measures show sensitivity to training load changes over time, although the strength and direction of associations between individual questionnaire items and training load are inconsistent. They function best as a complementary layer rather than a standalone monitoring tool and should be interpreted alongside objective data (Riboli et al., 2023).

Data quality depends heavily on athlete buy-in. When athletes perceive a lack of trust, unclear purpose, or fear of negative consequences, they may underreport fatigue or pain, undermining the entire monitoring process (Rebelo et al., 2026).

Designing an Integrated Monitoring Strategy

A monitoring system is not a single metric but a structured combination of tools that captures the training process from multiple angles. At minimum, an effective system should combine locomotor external load data with sRPE as an internal load indicator (Cormack & Coutts, 2022). This baseline captures both what was done and how it was perceived.

The MAA Framework

Tool selection should follow the Minimal, Adequate, and Accurate (MAA) framework: choose the fewest tools that provide adequate information with acceptable measurement accuracy (Rebelo et al., 2026). Adding more tools does not automatically improve decision-making. Each additional metric introduces data processing demands, personnel requirements, and interpretation complexity. The value of a tool lies not in how much data it generates, but in whether it changes a decision that would otherwise be made differently.

From Fatigue-Based to Training-Effects-Based Monitoring

Traditional monitoring paradigms focus on detecting fatigue. A more comprehensive approach frames monitoring around training effects — a concept that encompasses positive adaptation, maintenance, and maladaptation. Within this paradigm, readiness serves as an operational proxy for training effects: the absence of meaningful performance decrements, mental fatigue, or excessive psychological distress (Rebelo et al., 2026).

Readiness is not an independent outcome. It gains meaning only when interpreted alongside training load data, contextual information (travel, schedule density, life stressors), and longer-term trends. A single low CMJ score in isolation is ambiguous. A low CMJ score following three consecutive high-load days, combined with declining wellbeing scores, tells a coherent story that may warrant load modification.

The Quadrant Model for Decision-Making

Rebelo et al. (2026) proposed a quadrant model that plots two variables against each other to visualise the relationship between load and response: (A) training load × wellbeing, (B) training load × neuromuscular performance, and (C) neuromuscular performance × wellbeing. Each combination produces four quadrants representing different athlete states. An athlete in the high-load / low-wellbeing quadrant requires different management from one in the high-load / high-wellbeing quadrant. This visual approach supports rapid, intuitive decision-making without requiring complex statistical models.

Individualised Thresholds and Error Management

Meaningful change detection requires individual reference values. Establishing each athlete’s test-retest standard deviation (SD) enables the use of personalised thresholds: ±1 SD for sensitive detection, ±1.64 SD for 90% confidence, and ±2 SD for 95% confidence. In high-performance environments where the cost of missing an early sign of maladaptation is high (Type II error), more sensitive thresholds may be preferred (Rebelo et al., 2026).

Monitoring as a Decision-Support Tool

Monitoring does not replace professional judgement. It supplements it. The most sophisticated data collection system cannot account for every contextual factor that influences training decisions — tactical priorities, squad dynamics, fixture schedules, or an athlete’s personal circumstances. The practitioner’s role is to integrate quantitative monitoring data with qualitative observation and professional experience to arrive at decisions that serve the athlete’s development and availability (Rebelo et al., 2026).

Athlete buy-in is the prerequisite for data quality. Educating players on why data is collected, how it will be used, and what it will not be used for (e.g., punitive selection decisions) protects the integrity of both subjective and objective measures (Clubb & Murray, 2022).

---

Key Takeaways

• External load (prescribed physical work) and internal load (psychophysiological response during exercise) are distinct constructs; the internal response to identical external load varies between athletes depending on fitness, nutrition, psychological state, and genetics.
• Key external load tools include GPS/GNSS, optical tracking, and IMU/accelerometers; key internal load tools include HR-based methods (including TRIMP variants), RPE/sRPE, and wellbeing self-reports.
• Training adaptation is ultimately determined by internal load, so it should serve as the primary monitoring indicator while being integrated with external load to interpret fitness changes and fatigue status.
• “Training load” is a scientifically imprecise meta-construct under the SI system; practitioners should distinguish between volume and intensity or clearly specify the metric, tool, and construct being referenced.
• An effective monitoring system should at minimum combine locomotor external load with sRPE; athlete buy-in is a prerequisite for data quality, and monitoring supplements rather than replaces professional judgement.

References

1. Buchheit, M. & Simpson, B. M. (2017). Player-Tracking Technology: Half-Full or Half-Empty Glass?. International Journal of Sports Physiology and Performance, 12(s2), S2-35-S2-41. https://doi.org/10.1123/ijspp.2016-0499
2. Clubb, J., & Murray, A. M. (2022). Characteristics of tracking systems and load monitoring. In D. N. French & L. Torres Ronda (Eds.), NSCA’s Essentials of Sport Science. Human Kinetics.
3. Cormack, S., & Coutts, A. J. (2022). Training Load Model. In D. N. French & L. Torres Ronda (Eds.), NSCA’s Essentials of Sport Science. Human Kinetics.
4. Dello Iacono, A., Datson, N., Clubb, J., Lacome, M., Sullivan, A., & Shushan, T. (2025). Data analytics practices and reporting strategies in senior football: insights into athlete health and performance from over 200 practitioners worldwide. Science and Medicine in Football, 10(1), 80-95. https://doi.org/10.1080/24733938.2025.2476478
5. Impellizzeri, F. M., Marcora, S. M., & Coutts, A. J. (2019). Internal and External Training Load: 15 Years On. International Journal of Sports Physiology and Performance, 14(2), 270-273. https://doi.org/10.1123/ijspp.2018-0935
6. Jamieson, J. (2022). Heart rate and heart rate variability. In D. N. French & L. Torres Ronda (Eds.), NSCA’s Essentials of Sport Science. Human Kinetics.
7. Miguel, M., Oliveira, R., Loureiro, N., García-Rubio, J., & Ibáñez, S. J. (2021). Load Measures in Training/Match Monitoring in Soccer: A Systematic Review. International Journal of Environmental Research and Public Health, 18(5), 2721. https://doi.org/10.3390/ijerph18052721
8. Murray, A. M., & Clubb, J. (2022). Analysis of tracking systems and load monitoring. In D. N. French & L. Torres Ronda (Eds.), NSCA’s Essentials of Sport Science. Human Kinetics.
9. Pillitteri, G., Clemente, F. M., Sarmento, H., Figuereido, A., Rossi, A., Bongiovanni, T., Puleo, G., Petrucci, M., Foster, C., Battaglia, G., & Bianco, A. (2024). Translating player monitoring into training prescriptions: Real world soccer scenario and practical proposals. International Journal of Sports Science & Coaching, 20(1), 388-406. https://doi.org/10.1177/17479541241289080
10. Pimenta, R., Antunes, H., Silva, H., Ribeiro, J., & Nakamura, F. Y. (2026). The Need for GPS Data to be Normalized for Performance and Fatigue Monitoring in Soccer: Considerations for Accelerations and Decelerations. Strength & Conditioning Journal. https://doi.org/10.1519/ssc.0000000000000958
11. Rebelo, A., Bishop, C., Thorpe, R. T., Turner, A. N., & Gabbett, T. J. (2026). Monitoring training effects in athletes: A multidimensional framework for decision-making. Sports Medicine. Advance online publication. https://doi.org/10.1007/s40279-026-02417-4
12. Riboli, A., MacMillan, L., Calder, A., & Mason, L. (2023). Player monitoring and practical application. In A. Calder & A. Centofanti (Eds.), Peak performance for soccer: The elite coaching and training manual. Routledge.
13. Rice, J., Kovacevic, D., Calder, A., & Carter, J. (2023). Wearable technology. In A. Calder & A. Centofanti (Eds.), Peak performance for soccer: The elite coaching and training manual. Routledge.
14. Staunton, C. A., Abt, G., Weaving, D., & Wundersitz, D. W. (2022). Misuse of the term ‘load’ in sport and exercise science. Journal of Science and Medicine in Sport, 25(5), 439-444. https://doi.org/10.1016/j.jsams.2021.08.013