The aim of this study was to understand how endurance time, a proxy for physical fitness, has changed in healthy inner-city children and adolescents in the past three decades. This was a retrospective cross-sectional study. This study used exercise stress test in a laboratory of an inner-city teaching hospital. We reviewed all consecutive healthy children and adolescents who underwent an exercise Bruce protocol treadmill test from 1983 to 2010. The study population was divided into five groups of 5-year intervals based on the year of testing. Temporal trend in endurance time was analyzed, adjusting for gender, ethnicity, age, and body mass index (BMI). We analyzed the records of 435 healthy children and adolescents (mean age 12.6 ± 3.2 years, 57% males).There was a significant difference in the mean endurance time between children grouped in 5-year intervals (P < 0.001) with a significant downward trend in endurance time over the years (P < 0.001), especially after 2001. In contrast, there was no statistically significant change in the mean BMI between children grouped in 5-year intervals (P = 0.205). Multivariate linear regression model demonstrated that the date of testing was independently predictive of endurance time, adjusting for age, gender, BMI, and ethnicity (P < 0.001). Gender was the strongest independent predictor of endurance time, followed by age, BMI, and ethnicity. There is a downward trend in endurance time over the 27-year period among inner-city children and adolescents. Temporal decline in endurance time was independent of factors known to affect this parameter, such as age, gender, BMI, and ethnicity. Factors such as deconditioning due to sedentary lifestyle and lack of motivation to endure on the treadmill among later generations may have played a role in such decline.