Pregnant dams were fed a 20% protein diet with caffeine (2 mg/100 g b.wt.), starting on day 9 of gestation. At birth, each dam with 8 assigned pups was fed this diet until weaning, day 22. On day 22, female rats were caged and fed this diet until day 93. Starting on day 93, the caffeine-supplemented diet was replaced with a caffeine-free, 20% protein diet until day 388. Starting on day 31, each animal was placed in a photoactivity cage, and locomotive activity was measured until day 375. On day 388, the animals were killed, and their brains were removed and divided into 7 regions. The weight, DNA, protein and zinc contents, and alkaline phosphatase activity of each region were determined. Locomotive activity of the caffeine-fed group was higher than in the noncaffeine control group. Accumulative activity scores showed 3 subgroups (high, medium, and low) in both groups at day 93. The medium activity subgroup in the caffeine group was greater than the controls from day 72 to day 93. These differences reappeared 5 weeks after cessation of caffeine supplementation and continued until day 375. The differences in activity were minimum in the high and low subgroups. Chronic caffeine intake in early life permanently affected the medium activity subgroup. Furthermore, various regions of the brain were biochemically altered in spite of the feeding of a noncaffeine diet for almost 300 days after caffeine.