Intracranial pressure (ICP) monitoring procedure can be applied to aid in secondary brain damage prevention. A high invasiveness of commonly used ICP measuring methods poses a risk of complications, and therefore new non-invasive methods are currently being developed. A promising non-invasive ICP measurement method is based on the existence of pressure balance state, which is driven by the unique morphological property of ophthalmic artery (OA). The value of ICP can be obtained by evaluating blood flow or artery characteristics in different OA segments, intracranial OA segment (IOA) and extracranial OA segment (EOA). In order to increase measurement accuracy, the systematic errors must be evaluated, which requires an implementation of a numerical model encompassing various physical phenomena. In this paper, a developed numerical model is presented, which was used to solve a transient fluid–structure interaction (FSI) problem of the pulsatile blood flow in a straight, physically meaningful anisotropic, hyperelastic OA, with ICP and external pressure (Pe) loads. It was found that the systematic error based on mean cross-sectional area difference between IOA and EOA segments was {–1.48, –1.37, –1.17} mmHg with ICP = {10, 20, 30} mmHg, respectively. The systematic error based on mean blood flow velocity difference between IOA and EOA segments was {–1.84, –1.76, –1.625} mmHg with ICP = {10, 20, 30} mmHg, respectively. The presented numerical model examined the worst-case scenario in terms of boundary conditions, which were immovable, while lengths of OA segments were physiologically relevant statistical means; however, the obtained systematic errors still met the clinical standards of ANSI/AAMI, where it is stated that the error should not exceed the ± 2 mmHg in the range of 0–20 mmHg of ICP. Boundary conditions and compliance affects the systematic error in both ways (reduce or increase it); this may explain the low systematic errors obtained in experimental studies by other authors.