Objective <p>Understanding a patient’s nutrition literacy weaknesses may aid in improving the patient’s nutrition literacy and diet quality. The Nutrition Literacy Assessment Instrument (NLit) was developed to fill this knowledge gap. Global NLit cutoff scores were previously developed; however, subscale NLit cutoff scores are needed. </p> Design <p>In this secondary analysis, individual post hoc linear regressions were calculated for each NLit subscale to determine scoring thresholds for each NLit subscale. Scoring thresholds were determined for both long- and short-form NLit subscales.</p> Setting <p>Two University-affiliated general practice clinics and two University-affiliated safety net clinics were used as recruitment sites.</p> Participants <p>Patients (n = 402) with self-reported hyperlipidemia, hypertension, diabetes, or obesity were included in primary analysis from which data in this secondary analysis were derived.</p> Results <p>Subscale scores were associated with predicted Healthy Eating Index (HEI)-2010 scores linked to increased risk for mortality. Tenth-percentile NLit scores were associated with HEI-2010 scores between 59.33 and 62.27 (by subscale) for the long-form NLit, and between 60.12 and 61.77 for the short-form NLit.</p> Conclusions <p>NLit cutoff scores can pinpoint capacity with nutrition information by subscale. Identifying nutrition literacy deficits may allow nutrition professionals to better focus interventions.</p>