We measured the permeability of fluvastatin, lovastatin, and its pharmacologically active hydroxy acid (HA) across the blood-brain barrier (BBB) with three different techniques. The brain uptake index (BUI) technique yielded values for fluvastatin and HA below the limit of detection of this method; in contrast, brain extraction of lovastatin was high: 57 and 7% with addition of buffer and human plasma, respectively. Brain perfusion studies in rats indicated that when fluvastatin and HA were exposed for a longer time to the BBB, small but measurable brain uptake occurred, with permeability coefficients of 2.5 x 10(-4) and 1.4 x 10(-4) cm/min, for fluvastatin and HA, respectively, whereas the permeability coefficient of lovastatin was 2.3 x 10(-2) cm/min. An in vitro BBB model consisting of a primary culture of confluent monolayers of bovine brain microvessel endothelial cells yielded very similar values: 7.2 x 10(-4) cm/min for fluvastatin and 1.3 x 10(-3) cm/min for HA and a permeability coefficient of 1.1 x 10(-2) cm/min for lovastatin. The results of the different methods show that lovastatin crosses the BBB to a much greater extent than does fluvastatin or HA. Although lovastatin undergoes hydrolytic conversion to HA, it subsists in blood for sufficient time to cross the BBB. Therefore, lovastatin and fluvastatin are expected to be different in their central side effect profile, as was observed in patients treated for hypercholesterolemia. This finding supports the hypothesis that sleep disturbances due to some cholesterol biosynthesis inhibitors might be correlated with their abilities to cross the BBB.