Introduction

Training load refers to the input applied during training that elicits a physical and physiological response from the athlete (Impellizzeri et al., 2019). Training load can be described as prescribed (external) or actual (internal), depending on whether it refers to the work performed or the psychophysiological response elicited (Impellizzeri et al., 2019). The prescribed, or external, load is defined as the physical work specified in the training plan, including exercise selection, intensity, volume, and other loading variables (e.g., rest intervals between sets). The internal load refers to the psychophysiological response elicited within the athlete to cope with the demands imposed by the external load (Impellizzeri et al., 2019). The planned load corresponds to the intended internal load—that is, the level of stress targeted for the individual—yet it is operationally represented by the prescribed load.

During a training program, two of the main tasks of the coach are to: 1) ensure that the planned load is appropriate to produce the desired adaptations, and 2) ensure that the prescribed load accurately reflects the actual load. Failure to accurately align prescribed and actual loads may compromise progress within the training process, as the adaptations achieved will depend on the stimulus actually imposed rather than the one theoretically planned (Kraemer and Ratamess, 2004; Zatsiorsky et al., 2021). When the relationship among the external load (i.e., what is prescribed and performed), the internal load (i.e., the physiological and neuromuscular stress imposed), and the resulting adaptations is not objectively verified, the training process becomes a “black box”. Thus, objective monitoring strategies are essential to close this gap, ensuring a transparent and measurable link between prescription, execution, and chronic adaptations. To achieve this goal, it is essential to have a deep understanding of the two main components of training load: 1) intensity, usually described in relation to an objective or subjective maximum value of a given variable; and 2) volume, often expressed in units of time, distance, or the number of repetitions.

Training intensity is considered one of the most important variables when designing a resistance training (RT) program (Kraemer and Ratamess, 2004; Zatsiorsky et al., 2021). To individualize and facilitate programming, training intensity is commonly expressed relative to each athlete’s maximal capacity (Suchomel et al., 2018). In RT, relative intensity is typically expressed as a percentage of an individual’s one-repetition maximum (%1RM) (Cormie et al., 2011; Stone et al., 2007), that is, the maximum load an individual can lift during a maximal effort in a specific exercise (e.g., bench press or squat exercises). Nevertheless, the training effect of an RT session is determined by both the magnitude of relative intensity (Schoenfeld et al., 2017) and the total volume performed (sets × repetitions) (Rhea et al., 2003; Stone et al., 1999).

Specifically, training volume is intended to quantify the amount of work performed (Kraemer and Ratamess, 2004). This variable plays a central role in determining the responses and adaptations induced by RT (Grgic et al., 2022; Wernbom et al., 2007). However, as with intensity, the value of volume alone is insufficient to define the RT load, since the same volume yields a very different training stimulus when intensity varies. Several approaches have been used to determine RT volume. The most basic approach to quantifying training volume is the use of a repetition-based protocol, defined as the total number of repetitions performed during a given training session (Kraemer and Ratamess, 2004). While the simplicity of this procedure is advantageous for group settings, it is insufficient for accurately monitoring the internal training load experienced by each athlete. The main problem with this method is that it primarily considers external loads without providing accurate information about the fatigue they induce across athletes within a given training session (Halson, 2014; Impellizzeri et al., 2019). This complicates one of the coach’s main tasks: determining the actual load imposed on the athlete.

A recent consensus statement by the American College of Sports Medicine (ACSM) and Exercise & Sport Science Australia (ESSA) emphasizes the importance of considering progression toward neuromuscular failure when prescribing RT intensity, as fatigue accumulates with repetitions, especially near failure (Bishop et al., 2025). This consensus statement proposes the use of repetitions in reserve (RIR) as an alternative approach to training load prescription for both novice and experienced lifters (Bermúdez Droguett et al., 2025). The term of RIR was initially defined by Zourdos et al. (2016) as “the number of additional repetitions an individual believes they could complete before reaching muscular failure”. Despite the importance of considering proximity to task failure, the use of RIR as a standalone indicator presents several limitations that should be considered.

One key concern is the assumption that a given RIR value reflects a comparable training stimulus across different repetition ranges. This assumption implicitly treats proximity to failure as the sole determinant of exertion, disregarding the relative proportion of repetitions performed in relation to the total number of repetitions possible. This discrepancy may influence the magnitude of metabolic stress, neuromuscular fatigue, and subsequent recovery demands (González-Badillo and Sánchez-Medina, 2010; Sánchez-Medina and González-Badillo, 2011; Schoenfeld, 2010). For instance, equating RIR of 2 when only 2 repetitions are performed with RIR of 2 after 10 repetitions completed can be misleading: in the first case, the athlete completes 50% of the possible repetitions (2 of 4), whereas in the second, they complete approximately 83% (10 of 12). While both conditions technically result in a “two-repetition reserve”, the underlying physiological demands are quite distinct. Supporting this, Sánchez-Medina and González-Badillo (2011) reported marked differences in metabolic and mechanical stress during the squat exercise, including blood lactate (3.0 vs. 10.6 mmol•L⁻¹), blood ammonia (41 vs. 62 µmol•L⁻¹), jump height loss (5.7% vs. 16.6%), and strength impairment (6.1% vs. 14.6% for the 2 (of 4) and 10 (of 12) protocols, respectively).

In summary, these findings indicate that identical RIR values can correspond to different levels of accumulated fatigue and internal stress, thereby challenging the assumption that RIR alone can standardize the training stimulus across varying repetition configurations. From a programming perspective, this inconsistency may contribute to variability in fatigue accumulation and subsequent adaptive responses when similar RIR targets are prescribed under different loading schemes.

The Level of Effort Concept

The term “level of effort” was originally defined by González-Badillo and Gorostiaga (1995) and later empirically examined by Sánchez-Medina and González-Badillo (2011). It describes the work performed by an individual relative to their maximum strength capacity; that is, the relationship between the number of repetitions performed in a set and the total number of repetitions achievable (González-Badillo and Gorostiaga, 1995; Sánchez-Medina and González-Badillo, 2011). This concept expresses exertion proportionally, providing a more consistent representation of training demands based on the proportion of repetition capacity utilized within a set. Hence, the level of effort addresses the limitations of RIR by simultaneously considering two components: (a) the difference between repetitions performed and the maximum number of repetitions possible (i.e., RIR), and (b) the total repetition capacity associated with a given load. Accordingly, RIR represent only one component of the effort construct, as they reflect distance to failure without accounting for the relative workload completed within a respective set (González-Badillo et al., 2017).

Returning to the previous example, completing 2 (of 4) and 10 (of 12) repetitions yields the same difference of 2 repetitions between those performed and those possible, yet the acute physiological responses are considerably different. Although the level of proximity to failure is identical, the total repetition capacity differs substantially, thereby influencing acute fatigue, metabolic stress, and both central and peripheral fatigue responses (Sánchez-Medina and González-Badillo, 2011). This distinction is critical because physiological stress is not determined solely by how many repetitions remain, but by how much of the available repetition capacity has already been utilized. Therefore, a proportion-based representation of exertion more accurately reflects the internal load imposed during a set (Pareja-Blanco et al., 2017; Sánchez-Medina and González-Badillo, 2011) (Table 1).

In essence, the level of effort provides a more precise representation of the exertion imposed by a given load and better reflects the internal demands placed on the athlete. By integrating both the number of remaining repetitions and total repetition capacity, it establishes a standardized framework for interpreting training stress across varying repetition ranges and loading intensities, thereby overcoming the conceptual and structural limitations inherent in RIR when used in isolation.

Level of Effort Defined by Lifting Velocity

Currently, the level of effort can be accurately quantified through velocity-based measurements. The initial indicator of effort is determined by the difficulty of the first repetition; that is, by the relative loading intensity. Consequently, the level of effort is initially established by the velocity of the fastest repetition (typically the first or the second), which allows for a highly accurate estimation of the percentage of the 1RM (González-Badillo and Sánchez-Medina, 2010). To ensure a valid interpretation of effort, each repetition must be performed with maximal intent, as submaximal execution velocities fail to precisely reflect the true relative intensity of the load (Pareja-Blanco et al., 2014).

However, this alone does not fully define the level of effort, as overall exertion also depends on the proportion of the maximum number of possible repetitions performed: completing 1 repetition with a load that allows for 10 repetitions [1 (of 10)] is not equivalent to completing 8 repetitions with that same load [8 (of 10)]. Given that performing repetitions at maximal velocity induces a progressive decline in movement velocity, the level of effort is determined not only by the difficulty of the first repetition, but also by the velocity loss (VL) throughout the set (Sánchez-Medina and González-Badillo, 2011). As such, VL serves as an objective proxy for the proportion of repetition capacity utilized, providing a quantifiable expression of the proportion-based construct previously described (Bachero Mena et al., 2025; González-Badillo et al., 2017). The integration of fastest repetition velocity and intra-set VL represents a practical evolution of the level of effort construct (Pareja-Blanco et al., 2017, 2020).

By measuring VL within a set, it is also possible to estimate effort without knowledge of the total number of repetitions that could be performed, as strong relationships (R² = 0.92–0.97) have been observed between VL and the percentage of repetitions completed relative to the maximum achievable (%Rep, i.e., proximity to muscle failure) (González-Badillo et al., 2017; Sánchez-Moreno et al., 2021). Moreover, the %Rep completed at a given VL magnitude exhibits low variability (inter-individual coefficient of variation [CV]: 2.5–12.1%) and high reliability (intra-individual CV: 2.1–6.6%) (González-Badillo et al., 2017). Therefore, the %Rep corresponding to a given magnitude of VL appears to be relatively consistent across individuals, regardless of their maximum number of repetitions (González-Badillo et al., 2017; Sánchez-Moreno et al., 2021). This consistency supports the use of VL as a standardized indicator of intra-set effort (Bachero Mena et al., 2025; Rodiles-Guerrero et al., 2026). Conversely, the concept of RIR relies solely on the absolute number of repetitions remaining and does not account for differences in total repetition capacity (Helms et al., 2016; Zourdos et al., 2016).

This distinction is crucial for training prescription, as similar RIR values do not guarantee comparable physiological stress. It is important to emphasize that the present proposal does not aim to disregard the practical value of RIR in training prescription. On the contrary, RIR remain a widely adopted and useful method, and many RT programs can be designed with considerable precision using this strategy. Our intention is not to replace the RIR method, but rather to offer an alternative conceptual framework that enables coaches to define and monitor RT loads with greater proportional precision. By incorporating the level of the effort approach, strength and conditioning coaches may further enhance the individualization, effectiveness, and physiological precision of RT prescription.

Practical Implications

As a general guideline, and recognizing that effort exists on a continuum, the level of effort is considered low when the number of repetitions performed is far from the maximum achievable. In terms of intra-set VL, this typically corresponds to ~5–10% relative to the fastest repetition. Hence, this reflects a high initial velocity and minimal VL, with the repetitions performed generally representing less than half of the total capacity. As examples of low effort, the following values may be considered: 4–6 (of 16–30) or 3–4 (of 10–14). Although both scenarios are classified as “light”, their physiological impact differs considerably and would be prescribed in different contexts. From an applied perspective, interpreting effort relative to total capacity and VL may provide a more accurate reflection of the internal load than relying solely on RIR (Bastos et al., 2024; Mansfield et al., 2020).

The level of effort is considered moderate when approximately half of the possible repetitions are completed, corresponding to a VL of ~15–25%—magnitude associated with intermediate neuromuscular fatigue (Rodríguez-Rosell et al., 2020; Sánchez-Medina and González-Badillo, 2011); for example: 6–7 (of 12–14) or 4–5 (of 8–10) repetitions completed. Effort is categorized as high or very high when more than half of the possible repetitions are performed, implying a VL exceeding ~25–30%, yet with 2–4 repetitions remaining; for example: 3 (of 5), 4 (of 7), 5–6 (of 8), or 8 (of 12) repetitions completed. Effort is maximal when the maximum or near-maximum number of repetitions is completed, with VL reaching very high levels (50–70%), as illustrated by 9–10 (of 10), 7–8 (of 8), or 3–4 (of 4) repetitions completed. These categories are not rigid thresholds but practical reference points in which the magnitude of VL rather than the number of repetitions remaining dictates the internal stress imposed (Rodiles-Guerrero et al., 2026).

Nevertheless, these guidelines are general because, within this framework, when movement velocity is measured, the target is not a fixed number of repetitions, but rather the VL to be reached with a given load. This approach results in a different number of repetitions across individuals for the same level of effort. Such inter-individual variability is not a limitation but a core strength of this model, as it ensures exposure to equivalent levels of neuromuscular stress despite differences in repetition capacity, fiber composition, or fatigue resistance (Martinez-Canton et al., 2021; Pareja-Blanco et al., 2017). In contrast, prescribing identical RIR targets may lead to substantial discrepancies in accumulated fatigue and recovery demands. It is important to acknowledge that this is a conceptual article including an objective proposal, which should be tested in future studies comparing the two distinct methods (i.e., RT prescription based on RIR or level of effort approaches) and the potential neuromuscular adaptations provided by each of these methods. In addition, the dependence on linear position or velocity transducers to facilitate the implementation of the suggested approach may limit its application in less-equipped training facilities. Lastly, due to differences in training backgrounds and performance levels among athletes, the external validity of the proposed VL thresholds may be limited; therefore, these thresholds might need to be adjusted on an individual basis.

Conclusions

In conclusion, relying solely on RIR may provide an incomplete representation of the actual exertion imposed during RT, as it disregards the total number of possible repetitions, which fundamentally modulates the physiological stimulus. The level of effort approach addresses this limitation by integrating both components of exertion—repetitions performed and repetitions possible—thereby providing a clearer and more precise understanding of the internal demands placed on the athlete. By adopting a proportion-based perspective, this construct overcomes the structural limitations inherent in methods that rely exclusively on the absolute number of repetitions remaining. This concept enhances the interpretation of proximity to failure and improves the standardization of training stress across diverse repetition ranges, loading schemes, and individual capacities. When available, velocity-based feedback serves as a robust tool to objectively quantify exertion, providing an empirically supported and low-variability indicator of the proportion of repetition capacity utilized. In applied training settings, this approach allows practitioners to derive RIR from an objective quantification of effort, rather than relying solely on RIR to infer exertion. As a construct, the level of effort represents a more comprehensive and informative metric than RIR alone. Adopting this perspective as a core principle in RT prescription enhances both scientific rigor and practical decision-making, enabling coaches and sport scientists to better align prescribed training with the intended adaptive stimulus. The level of effort provides a coherent, standardized, and physiologically grounded framework for prescribing and monitoring RT load.