Boronophenylalanine (BPA) is used as Boron-10 carrier in boron neutron capture therapy, an experimental cancer radiotherapy. Results of quantitative, noninvasive in vivo detection and imaging of BPA in laboratory animals using 1 H NMR are presented for the first time. The purpose of this study was to implement and validate optimized techniques for the efficient detection of BPA. The 1 H NMR signals through which BPA is most readily detected in vivo are those from the aromatic ring of the molecule, which are part of a scalar-coupled spin system. The preferred detection method should therefore be based on a pulse sequence in which the effective TE is as short as possible. Modified versions of LASER ( CP ؍ 4.6 ms, TE ؍ 27.6 ms) and double-echo slice-selective 2D MRSI (TE ؍ 12 ms) were implemented for single-voxel spectroscopy and spectroscopic imaging of BPA, respectively. Chemical shift selective excitation was used for both sequences, based on a pulse that enabled narrow-band excitation without concomitant delay in TE. SI data without water suppression was used for absolute quantitation and for correction of Boron neutron capture therapy (BNCT) is a binary cancer treatment, in which 10 B nuclei, preferably targeted to tumors by suitable carrier molecules, are irradiated with low-energy neutrons, causing localized short-range and cell-damaging radiation (1). One of the problems confronting successful clinical implementation of BNCT is the difficulty of quantitatively mapping the distribution of the boronated (and 10 B-enriched) molecules in patients in the course of the treatment session. So far, MR is the only modality capable of fulfilling this task. Detection of the NMR-sensitive 10 B spin is the most straightforward approach to this problem, which was shown to be feasible for BSH, another molecule used in BNCT (2). However, in boronophenylalanine (BPA) the 10 B transverse relaxation is far more rapid than in BSH, so that direct 10 B detection is at much greater disadvantage compared to 1 H detection.The possibility of detecting the aromatic BPA protons was recognized previously and demonstrated on a patient using standard STEAM (TE ϭ 30 ms) (3). Subsequent studies characterized the echo-time dependence of the aromatic proton BPA signals in aqueous solutions for STEAM and PRESS sequences (4,5). As expected, these studies demonstrated the considerable modulation of the signal due to J-coupling, causing both phasing and signal loss problems and stressing the need for efficient, short-TE sequences. Echo modulation by J-coupling can be suppressed by using very short echo times or CP-type echo trains or by timing the detection to peaks of this modulation. The latter approach is also useful for spectral editing, but for 1 H couplings with low J values, this leads to rather long echo times and signal loss from intrinsic T 2 relaxation. The use of short-TE localized MRS or SI was demonstrated for both STEAM-(6) and PRESS-based sequences (7). Sequences detecting the full Hahn spin echo have, in principle, a twofol...