Purpose To evaluate agreement of total corneal power (TCP) measured by swept-source anterior segment optical coherence tomography (CASIA 2), Scheimpflug tomography (Pentacam AXL), and swept-source optical biometer (IOLMaster 700) in normal and keratoconic patients. Methods This is a prospective observational study conducted at King Chulalongkorn Memorial Hospital, Bangkok, Thailand. Biometric values were measured by each device for three times by two operators to evaluate repeatability and reproducibility of TCP. The agreement of TCP and other parameters including total corneal astigmatism, anterior keratometry, anterior corneal astigmatism, posterior keratometry, posterior corneal astigmatism, anterior chamber depth, white-to-white corneal diameter (WTW), central corneal thickness, and intraocular power were also evaluated. Results This study enrolled 100 healthy participants and 34 patients with keratoconus. The repeatability and reproducibility of TCP were high in all devices (ICC > 0.9). The agreement of TCP was excellent among three devices in both groups (ICC > 0.9). However, the agreement of TCP between CASIA 2 and IOLMaster 700 was slightly lower in healthy participants (ICC = 0.867) and keratoconic patients (ICC = 0.852) with mean differences of more than 1.0 diopter is clinically significant. Wider 95% limit of agreement were found in keratoconic eyes. Most of other parameters showed good to excellent agreement except WTW which showed poor to moderate agreement in both groups. IOL power showed clinically significant differences in patients with keratoconus. Conclusions The agreement of TCP measured by three devices was excellent in normal and keratoconic patients. However, TCP cannot be used interchangeably between devices.