The use of prostate-specific antigen (PSA) as a screening test remains controversial. There have been several attempts to refine PSA measurements to improve its predictive value. These modifications, including PSA density, PSA kinetics, and the measurement of PSA isoforms, have met with limited success. Therefore, complex statistical and computational models have been created to assess an individual's risk of prostate cancer more accurately. In this review, the authors examined the methods used to modify PSA as well as various predictive models used in prostate cancer detection. They described the mathematical underpinnings of these techniques along with their intrinsic strengths and weaknesses, and they assessed the accuracy of these methods, which have been shown to be better than physicians' judgment at predicting a man's risk of cancer. Without understanding the design and limitations of these methods, they can be applied inappropriately, leading to incorrect conclusions. These models are important components in counseling patients on their risk of prostate cancer and also help in the design of clinical trials by stratifying patients into different risk categories. Thus, it is incumbent on both clinicians and researchers to become familiar with these tools. Despite intensive work over the last several decades, prostate cancer continues to be the most common cancer in men and their second leading cause of cancer death. In 2008, it is estimated that 186,320 men received a new diagnosis of this disease, and nearly 29,000 died from it.1 However, there is no universal agreement on screening for prostate cancer. The American Urological Association and the American Cancer Society both recommend using a combination of digital rectal examination (DRE) and serum prostatespecific antigen (PSA) level to screen low-risk white men starting at age 50 years and to screen high-risk populations (including African Americans) starting at age 40 years. Much of the dilemma has to do with limitations of the PSA test itself, which has poor sensitivity and specificity. Various attempts to improve the predictive capacity of PSA likewise have met with only partial success. Because of these deficiencies, statistical and computational models have been created to more accurately predict a patient's risk of prostate cancer at biopsy. This, in turn, helps patients make a more informed decision concerning the choice to proceed with biopsy, a decision that should not be made lightly given the costs involved, the possibility of complications, and the chance of diagnosing clinically insignificant cancer. In this review, we provide a brief summary of PSA and its permutations and examine the methods that generate the predictive models as well as their inherent strengths and weaknesses.Overview of PSA and PSA-specific antigen modifications Catalona et al initially explored the use of PSA as a screening tool. In 2 large studies of healthy men aged !50 years, a higher screening PSA level was correlated with a greater likelihood of cancer. 5,6 Furth...