MRI of the lungs using hyperpolarized helium-3 ( 3 He) allows the determination of intrapulmonary oxygen partial pressures (p O2 ). The need to separate competing processes of signal loss has hitherto required two different imaging series during two different breathing maneuvers. In this work, a new imaging strategy to measure p O2 by a single series of consecutive scans is presented. Within the last 5 years, helium-3 magnetic resonance imaging ( 3 He-MRI) of the lungs has been employed for both morphological and functional imaging. The first encompasses studies of overt or subclinical lung disease, e.g., in chronic obstructive pulmonary disease (COPD) and emphysema patients (1,2), asthma patients (3), and smokers (4). The latter includes the use of fast imaging sequences (5-7), diffusion studies (8 -10), and the determination of intrapulmonary oxygen partial pressures (11-13). All of these methods provide exciting new approaches to lung function analysis. Complementary to the diagnostic information obtained from morphological imaging, rapid lung imaging allows time-resolved ventilation studies. Diffusion-weighted imaging and oxygen-sensitive MRI yield physiological information that was previously unavailable or was obtainable only by invasive means. (Comprehensive reviews are given in Refs. 14 and 15.)Oxygen-sensitive 3 He-MRI makes use of the oxygen-induced nuclear relaxation of 3 He (16) to compute the intrapulmonary oxygen partial pressure p O2 during short breath-holds. However, true relaxation due to molecular oxygen must be distinguished from RF-induced signal loss, i.e., the two depolarization effects must be decoupled. As neither is known a priori, this requires two different imaging sequences: with different RF excitation amplitudes and, hence, flip angles (11), or with different interimage time intervals (12). A subtraction of the logarithmic intensities of both image series serves to mathematically eliminate one unknown and allows precise determination of the other. To date, the two image series have been acquired during two separate breathing maneuvers.The analyses described in Refs. 11 and 12 rely on identical physiological conditions during paired image acquisition. It has to be assumed that the p O2 at the start of the imaging sequence and its temporal development are equal in both breath-holds. The imaged subject thus has to perform two identical breathing maneuvers. However, preliminary studies in healthy human volunteers (13) and patients (17) have shown that this condition is not always met.A second disadvantage of the double-acquisition approach lies in its 3 He consumption. In previous experiments (12,13), a 3 He amount of ϳ2 ϫ 200 cm 3 was required for each examination. In view of the limited resources of 3 He, a more economical measurement technique is desirable.In this work, we present a new imaging algorithm which permits the determination of intrapulmonary p O2 using a single-breath, single-bolus imaging sequence. Its strategy is to employ a simultaneous change in both excitation ampli...