Background and Objective. Plasma phosphorylated tau (p-tau) has emerged as a promising biomarker for Alzheimer's disease (AD). Studies have reported strong associations between p-tau and tau-PET that are mainly driven by differences between amyloid-positive and amyloid-negative patients. However, the relationship between p-tau and tau-PET is less characterized within cognitively impaired patients with a biomarker-supported diagnosis of AD. We conducted a head-to-head comparison between plasma p-tau217 and tau-PET in patients at the clinical stage of AD and further assessed their relationships with demographic, clinical, and biomarker variables. Methods. We retrospectively included 87 amyloid-positive patients diagnosed with MCI or AD dementia who underwent structural MRI, amyloid-PET (11C-PIB), tau-PET (18F-flortaucipir, FTP), and blood draw assessments within one year (age=66 ±10, 48% female). Amyloid-PET was quantified in Centiloids (CL) while cortical tau-PET binding was measured using Standardized Uptake Value Ratios (SUVRs) referenced against inferior cerebellar cortex. Plasma p-tau217 concentrations were measured using an electrochemiluminescence-based assay on the Meso Scale Discovery platform. MRI-derived cortical volume was quantified with FreeSurfer. Mini-Mental State Examination (MMSE) scores were available at baseline (n=85) and follow-up visits (n=28; 1.5 ± 0.7 years). Results. Plasma p-tau217 and cortical FTP-SUVR were correlated (r=0.61, p<.001), especially in temporo-parietal and dorsolateral frontal cortices. Both higher p-tau217 and FTP-SUVR values were associated with younger age, female sex, and lower cortical volume, but not with APOE-ϵ4 carriership. PIB-PET centiloids were weakly correlated with FTP-SUVR (r=0.26, p=0.02), but not with p-tau217 (r=0.10, p=0.36). Regional PET-plasma associations varied with amyloid burden, with p-tau217 being more strongly associated with tau-PET in temporal cortex among patients with moderate amyloid-PET burden, and with tau-PET in primary cortices among patients with high amyloid-PET burden. Higher p-tau217 and FTP-SUVR values were independently associated with lower MMSE scores cross-sectionally, while only baseline FTP-SUVR predicted longitudinal MMSE decline when both biomarkers were included in the same model. Discussion. Plasma p-tau217 and tau-PET are strongly correlated in amyloid-PET-positive patients with MCI and AD, and they exhibited comparable patterns of associations with demographic variables and with markers of downstream neurodegeneration.