The development of noninvasive techniques should always be an important goal of diagnostic radiology. Despite the recent excitement regarding intravenous MR imaging contrast agents (e.g., Gd-DTPA), the role of noncontrast MR imaging for the breast should not be overlooked. First, it must be kept in mind that administration of an intravenous contrast material is an invasive procedure that poses some health concern and may not be acceptable to certain patients. Second, injection of intravenous contrast media is an expensive procedure that incurs costs related to the contrast media itself as well as from the personnel required to administer the media and monitor any adverse reactions. Third, there may be specific imaging situations in which a lesion must be located prior to administration of contrast media, for example, when dynamic contrast enhanced imaging is to be performed and temporal resolution limits the volume of coverage to only a few slices. Finally, noncontrast MR imaging techniques also may be useful in their own right for obtaining information not otherwise available from a contrast-enhanced study, particularly quantitative relaxometric and spectroscopic measurements that may be useful for tissue characterization. It should also be noted that optimized noncontrast MR methods also might be used in conjunction with contrast agents to improve overall clinical utility. The heterogeneous nature of the breast and the overlap in T1 and T2 between different normal and abnormal breast tissue types frequently renders conventional noncontrast breast MR images confusing and ambiguous. Hybrid imaging approaches combining T1 weighting, T2 weighting, and fat suppression techniques as well as multivariate image analysis may be helpful for improving the sensitivity of MR imaging to breast disease; however, specificity remains a problem. Recent advances in clinical MR imaging have provided the opportunity to explore other approaches which may shed new light on the characterization of breast disease including NMRD, spin locking, magnetization transfer, and multicomponent relaxation analysis. MR imaging derived morphologic factors (i.e., lesion spiculation, parenchymal patterns) also may have a role to play in breast disease diagnosis and risk assessment. Several avenues of future development of noncontrast breast MR imaging would seem fruitful. NMRD and spin locking results suggest the potential contrast benefit of low field imaging of the breast, particularly, with the probable development of open concept low field MR imaging devices, which will provide improved access and low cost. In vivo relaxation time measurements, particularly multi-component T2 analysis and incorporation of MTC and chemical shift imaging techniques, may improve breast MR image specificity.(ABSTRACT TRUNCATED AT 400 WORDS)