Introduction Fluconazole is an important antifungal in the prevention and treatment of invasive Candida infections in neonates, even though its use in preterm infants is still off-label. Here, we performed a population pharmacokinetic study on fluconazole in preterm neonates in order to optimise dosing through the identified predictive patient characteristics. Methods Fluconazole concentrations obtained from preterm infants from two studies were pooled and analysed using NONMEM V.7.3. The developed model was used to evaluate current dosing practice. A therapeutic dosing strategy aiming to reach a minimum target exposure of 400 and 200 mg×hour/L per 24 hours for fluconazole-susceptible C. albicans meningitis and other systemic infections, respectively, was developed. Results In 41 preterm neonates with median (range) gestational age 25.3 (24.0–35.1) weeks and median postnatal age (PNA) at treatment initiation 1.4 (0.2–32.5) days, 146 plasma samples were collected. A one-compartment model described the data best, with an estimated clearance of 0.0147 L/hour for a typical infant of 0.87 kg with a serum creatinine concentration of 60 μmol/L and volume of distribution of 0.844 L. Clearance was found to increase with 16% per 100 g increase in actual body weight, and to decrease with 12% per 10 μmol/L increase in creatinine concentration once PNA was above 1 week. Dose adjustments based on serum creatinine and daily dosing are required for therapeutic target attainment. Conclusion In preterm neonates, fluconazole clearance is best predicted by actual body weight and serum creatinine concentration. Therefore, fluconazole dosing should not only be based on body weight but also on creatinine concentration to achieve optimal exposure in all infants.