BACKGROUND Although mass screening for prostate cancer does not meet the criteria for an effective screening programme, multiphasic screening which includes PSA testing is still being carried out. OBJECTIVE We decided to study and evaluate the usefulness of PSA testing in multiphasic health screening and at the same time establish age-specific ranges of normal PSA values in our local population. RESULTS Six hundred and ninety five male patients who had their PSA levels tested during a multiphasic health screening from October 1992 to August 1995 were evaluated. Abnormal PSA levels were repeated and subjected to a DRE and TRUS biopsy if they were persistently high using age-specific PSA ranges. Our results showed 14 (4.1%) out of 695 patients who had an abnormal PSA of > 4 ng/mL. compared to 19 who had abnormal PSA levels using the age-specific PSA ranges. Of the patients who were < 40 yrs of age, no further investigations were done. Amongst those 80 years and older, none had abnormal age-specific PSA rates. No prostate cancers were picked up amongst all the patients investigated. Median age specific PSA values at the 95th percentile was calculated for each age group. A rise in the median PSA values with age was also noted. CONCLUSIONS We recommend that in patients less than 40 years of age, PSA should not be carried out as the probability of prostate cancer is almost zero. Similarly, in patients who are 80 years and above and asymptomatic, such screening may not be indicated given the limited options available. Age-specific rates are a better way to reduce the negative biopsy rates in the age-groups that are amenable to curative treatment. With a local set of age-specific PSA ranges, we hope to increase the positive predictive value of PSA for prostate cancers in our local population until more specific and equally sensitive tests are made available.