The purpose of the study was to: 1) identify the differences in serum testosterone levels among four groups of adult men differentiated on the basis of physical fitness and age, and 2) determine the effect of a four-month physical fitness programme consisting of running, calisthenics and recreational activities on the serum testosterone levels of the four groups. The groups were designated: high-fit, young aged about 32 (n = 7); high-fit, old aged about 52 (n = 7); low-fit, young (n = 7), and low-fit, old (n = 7). The subjects were selected and grouped according to physical fitness scores obtained using the regression equation of Ismail et al. Serum testosterone was determined by a radioimmunoassay method. The pre-test ANOVA revealed that the high-fit groups had a significantly (p less than .01) higher testosterone level (754.29 ng/100 ml) than the low-fit groups (548.07 ng/100 ml) and the high-fit, young group (925.01 ng/100 ml) was significantly (p less than .01) higher than the other three groups. Post-test values were adjusted using pre-test testosterone values as covariates. No significant differences among the groups were found indicating that the serum testosterone levels were the same regardless of different ages and fitness levels. The findings were discussed in light of physiological, biochemical and psychological factors.