The GRADE system classifies recommendations in guidelines as strong or weak. Strong recommendations indicate that the benefits of an intervention outweigh the risks, while weak recommendations suggest that the benefits are likely to outweigh the risks, but with less confidence. The strength of a recommendation is determined by four factors: the balance of benefits and harms, the quality of evidence, variability in patient values and preferences, and resource use. Strong recommendations mean that most informed patients would choose the recommended course of action, and clinicians can structure their interactions accordingly. Weak recommendations indicate that patients' choices may vary based on their values and preferences, and clinicians must ensure that care aligns with these values. The balance between desirable and undesirable consequences of alternative management strategies is the first determinant of recommendation strength. For example, antenatal steroids for premature births are strongly recommended because their benefits far outweigh the risks. In contrast, for low-risk atrial fibrillation patients, the choice between warfarin and no treatment is weak because the risks and benefits are closely balanced. The quality of evidence is the second determinant. If the evidence is uncertain, a strong recommendation may not be appropriate. For example, graduated compression stockings may have a large effect on reducing deep venous thrombosis, but flawed studies mean a weak recommendation is given. The third determinant is variability in patient values and preferences. For instance, in stroke prevention for atrial fibrillation, most patients are willing to accept a bleeding risk to avoid a stroke, but some are not. This variability affects the strength of the recommendation. The fourth determinant is cost. Cost varies widely, and recommendations must consider the context in which they are applied. Strong recommendations may not be important from all perspectives. If the consequences of a choice are relatively unimportant, some patients may not follow even strong recommendations. Governments and public health officials must also consider factors beyond the strength of a recommendation, such as the prevalence of the health problem and equity. Recommendations for use in research may be appropriate when there is insufficient evidence of benefit. Such recommendations can encourage further research. The GRADE system uses symbolic representations for quality of evidence and strength of recommendations, which are preferred for clarity. The system is used by many guideline panels, and its principles are widely accepted. The GRADE Working Group includes many experts in clinical epidemiology and health services research. The system has been peer-reviewed and is supported by various organizations. The system is used to improve the quality of healthcare guidelines and decision-making.The GRADE system classifies recommendations in guidelines as strong or weak. Strong recommendations indicate that the benefits of an intervention outweigh the risks, while weak recommendations suggest that the benefits are likely to outweigh the risks, but with less confidence. The strength of a recommendation is determined by four factors: the balance of benefits and harms, the quality of evidence, variability in patient values and preferences, and resource use. Strong recommendations mean that most informed patients would choose the recommended course of action, and clinicians can structure their interactions accordingly. Weak recommendations indicate that patients' choices may vary based on their values and preferences, and clinicians must ensure that care aligns with these values. The balance between desirable and undesirable consequences of alternative management strategies is the first determinant of recommendation strength. For example, antenatal steroids for premature births are strongly recommended because their benefits far outweigh the risks. In contrast, for low-risk atrial fibrillation patients, the choice between warfarin and no treatment is weak because the risks and benefits are closely balanced. The quality of evidence is the second determinant. If the evidence is uncertain, a strong recommendation may not be appropriate. For example, graduated compression stockings may have a large effect on reducing deep venous thrombosis, but flawed studies mean a weak recommendation is given. The third determinant is variability in patient values and preferences. For instance, in stroke prevention for atrial fibrillation, most patients are willing to accept a bleeding risk to avoid a stroke, but some are not. This variability affects the strength of the recommendation. The fourth determinant is cost. Cost varies widely, and recommendations must consider the context in which they are applied. Strong recommendations may not be important from all perspectives. If the consequences of a choice are relatively unimportant, some patients may not follow even strong recommendations. Governments and public health officials must also consider factors beyond the strength of a recommendation, such as the prevalence of the health problem and equity. Recommendations for use in research may be appropriate when there is insufficient evidence of benefit. Such recommendations can encourage further research. The GRADE system uses symbolic representations for quality of evidence and strength of recommendations, which are preferred for clarity. The system is used by many guideline panels, and its principles are widely accepted. The GRADE Working Group includes many experts in clinical epidemiology and health services research. The system has been peer-reviewed and is supported by various organizations. The system is used to improve the quality of healthcare guidelines and decision-making.