BACKGROUND: Diagnostic tools for familial hypercholesterolemia (FH) rely on estimation of LDL cholesterol concentration. However, routine measurement or calculation of LDL cholesterol concentration using the Friedewald equation contains a cholesterol contribution from lipoprotein(a) [Lp(a)]. We investigated whether Lp(a) influences the phenotypic diagnosis of FH by commonly used clinical criteria. METHODS: A cohort of 907 adult index patients attending a clinic were studied. The Dutch Lipid Clinic Network (DLCN) and Simon Broome (SB) diagnostic criteria were estimated before and after adjusting LDL cholesterol concentration for the cholesterol content (30%) of Lp(a). Diagnostic reclassification rates and area under the ROC (AUROC) curves in predicting an FH mutation were also compared. RESULTS: Seventy-four patients defined by DLCN criteria (8.2%) and 207 patients defined by SB criteria (22.8%) were reclassified to "unlikely" FH after adjusting LDL cholesterol for Lp(a) cholesterol. The proportion of FH patients defined by DLCN (probable/definite) and SB (possible/definite) criteria decreased significantly in patients with increased Lp(a) (>0.5 g/L; n = 330) after Lp(a) cholesterol adjustment (P < 0.01). The overall reclassification rate was significantly higher in patients with Lp(a) concentration >1.0 g/L (P < 0.001). The AUROC curve for LDL cholesterol concentration ≥191 mg/dL (≥5.0 mmol/L), DLCN criteria, and SB criteria in predicting an FH mutation increased significantly after adjustment (P < 0.001). There was no significant difference in AUROC curve before and after Lp(a) cholesterol adjustment at an LDL cholesterol concentration ≥251 mg/dL (≥6.5 mmol/L). CONCLUSIONS: Adjusting LDL cholesterol concentration for Lp(a) cholesterol improves the diagnostic accuracy of DLCN and SB criteria, especially with Lp(a) >1.0 g/L and LDL cholesterol <251 mg/dL (<6.5 mmol/L). Lp(a) should be measured in all patients suspected of having FH. © 2019 American Association for Clinical Chemistry.