Luteal phase deficiency is an ovulatory dysfunction problem that is subtle but real. It may be the most common ovulatory problem in women. Luteal phase deficiency has been clearly demonstrated in the research setting (1) in spontaneous cycles, (2) when follicular maturation has been impeded, and (3) when luteotrophic influences have been suppressed. The diagnosis of LPD in the clinical setting remains problematic and controversial primarily because there is no practical diagnostic method that has been validated. This article has reviewed the methods that have been used to diagnose LPD. BBT charts are insensitive; these charts reliably diagnose LPD only when there are persistent short luteal phases. There is disagreement whether ovarian follicular size, as determined by ultrasonography, is decreased in LPD; however, ultrasonographic diagnosis of LPD would require daily scans through ovulation, which makes this approach impractical. Mild hyperprolactinemia is a probable cause of LPD in a minority of patients; a physician should obtain a PRL level in LPD women with the realization that there is considerable sampling variability. Determination of serum gonadotropin levels (LH or FSH or both) is not practical for the clinical diagnosis of LPD. Random serum P levels, whether single or multiple, are not helpful in the diagnosis of LPD in individual patients. The secretory pattern of P results in such wide confidence limits that P samples from individuals cannot be compared to normal in a useful manner. Most of the controversy about the diagnosis of LPD has centered around the use of individual serum P levels. The timed endometrial biopsy relies on the endometrium as a bioassay of P over time. The endometrial biopsy has not been carefully validated in terms of its sensitivity or accuracy for the diagnosis of LPD. However, it remains the best current method for the diagnosis of LPD when the standard guidelines for its use are followed. As opposed to the other tests for LPD, awareness of the usefulness of the biopsy has increased as we have learned more about CL physiology. No current research method for the diagnosis of LPD appears to be a practical method that could be applied in the clinical setting. Specific secretory proteins from the endometrium and methods to measure hormone secretion that circumvent the secretory pattern hold promise for improved methods to diagnose LPD in the future.