Laminates of Bismaleimide (BMI) resin with Quartz (AQ581) fiber reinforcement are frequently used in aerospace applications, such as engine cowlings, radomes, and similar structural components. Accurate predictive models for absorption of liquid penetrants are particularly important for these composite components, which are often susceptible to long-term degradation due to absorbed moisture, hydraulic fluids, or similar penetrants. Although the microstructural features such as fiber volume fraction and void content of composite laminates are known to have a significant effect on absorption, accurate predictive models for non-Fickian absorption behavior supported by experimental data have not been available for Quartz/BMI laminates. In this paper, hydraulic fluid absorption characteristics of Quartz/BMI laminates manufactured by prepregs conditioned at different relative humidity and subsequently cured at different pressures are presented. The composite samples are immersed into hydraulic fluid at room temperature, and were not subjected to any prior degradation. To generate different levels of process-induced microvoids, prepregs were conditioned in an environmental chamber at 2% or 99% relative humidity at room temperature for a period of 24 hours prior to laminate fabrication. To alter the fiber volume fraction, the laminates were manufactured at cure pressures of 10 or 70 psi via a hot-press. The laminates are shown to have different levels of microvoids and fiber volume fractions, which are also observed to affect the absorption dynamics considerably and exhibited clear non-Fickian behavior. A one-dimensional Hindered Diffusion Model (HDM) was shown to be successful in predicting the hydraulic fluid absorption, including short-term diffusion and long-term anomalous behavior. Model prediction indicates that as the manufacturing pressure increased from 10 psi to 70 psi, the maximum absorbed fluid content (M∞) decreases from 7.5% wt. to 0.9% wt. In addition, the degree of non-Fickian behavior, measured by hindrance coefficient (µ), is shown to increase due to the process-induced microvoids.