Background and Objective For patients with intracranial infection, local intrathecal administration of meropenem may be a useful method to obtain a sufficient drug concentration in the cerebral spinal fluid (CSF). However, a large inter-individual variability may pose treatment efficacy at risk. This study aimed to identify factors affecting drug concentration in the CSF using population pharmacokinetics method. Methods After craniotomy, aneurysm patients with an indwelling lumbar cistern drainage tube who received a combined intravenous and intrathecal administration of meropenem for the treatment of suspected intracranial infection were enrolled. Venous blood and CSF specimens were collected for determining meropenem concentrations. Nonlinear mixed-effects modeling method was used to fit blood and CSF concentrations simultaneously and to develop the population pharmacokinetic model. The proposed model was applied to simulate dosage regimens. Results A three-compartmental model was established to describe meropenem in vivo behavior. Lumbar CSF drainage resulted in a drug loss, and drug clearance in CSF (CLCSF) was employed to describe this. The covariate analysis found that the drainage volume (mL/day) was strongly associated with CLCSF, and the effect of creatinine clearance was significant on the clearance of meropenem in blood (CL). Visual predictive check suggested that the proposed pharmacokinetic model agreed well with the observations. Simulation showed that both intravenous and intrathecal doses should be increased with the increases of minimum inhibitory concentration and daily CSF drainage volume. Conclusion This model incorporates covariates of the creatinine clearance and the drainage volume, and a simple to use dosage regimen table was created to guide clinicians with meropenem dosing.