The purpose of this study was to determine the immediate effects of cognitive training on healthy older adults and verify the transfer effects of targeted and non-targeted abilities. The design consisted of a semi-randomized clinical controlled trial. The final sample was composed of 80 volunteers recruited from a Brazilian community (mean age = 69.69; SD = 7.44), which were separated into an intervention group (N = 47; mean age = 69.66, SD = 7.51) and a control group (N = 33; mean age = 69.73, SD = 7.45). Intervention was characterized by adaptive cognitive training with 12 individual training sessions of 60 to 90 minutes (once a week). Eight instruments were used to assess effects of cognitive training. Five were used to assess trained abilities (near effects), including: Memorization Tests (List and History), Picture Completion, Digit Span, Digit Symbol-Coding, and Symbol Search (the last four from WAIS-III). Two instruments assessed untrained abilities (far effects): Arithmetic and Matrix Reasoning (WAIS-III). The non-parametric repeated measures ANOVA test revealed a significant interaction between group by time interaction for Picture Completion [F(74) = 14.88, p = .0002, d = 0.90, CLES = 73.69%], Digit Symbol-Coding [F(74) = 5.66, p = .019, d = 0.55, CLES = 65.21%] and Digit Span [F(74) = 5.38, p = .02, d = 0.54, CLES = 64.85%], suggesting an interventional impact on these performance tasks. The results supported near transfer effects, but did not demonstrate a far transfer effects.