Cognitive abilities of children in Costa Rica and Austria were compared using three age groups (N = 385/366). Cognitive ability tests (mental speed, culture reduced/fluid intelligence, literacy/crystallized intelligence) were applied that differed in the extent to which they refer to school-related knowledge. Preschool children (kindergarten, 5-6 years old, N(CR) = 80, N(Au) = 51) were assessed with the Coloured Progressive Matrices (CPM), primary school children (4th grade, 9-11 years old, N(CR) = 71, N(Au) = 71) with ZVT (a trail-making test), Standard Progressive Matrices (SPM) and items from PIRLS-Reading and TIMSS-Mathematics, and secondary school students (15-16 years old, N(CR) = 48, N(Au) = 48) with ZVT, Advanced Progressive Matrices (APM) and items from PISA-Reading and PISA-Mathematics. Additionally, parents and pupils were given questionnaires covering family characteristics and instruction. Average cognitive abilities were higher in Austria (Greenwich-IQ M(CR) = 87 and M(Au) = 99, d(IQ) = 12 points) and differences were smaller in preschool than in secondary school (d(IQ) = 7 vs 20 points). Differences in crystallized intelligence were larger than in fluid intelligence (mental speed: d(IQ) = 12, Raven: d(IQ) = 10, student achievement tests: d(IQ) = 17 IQ points). Differences were larger in comparisons at the level of g-factors. Austrian children were also taller (6.80 cm, d = 1.07 SD), but had lower body mass index (BMI(CR) = 19.35 vs BMI(Au) = 17.59, d = -0.89 SD). Different causal hypotheses explaining these differences are compared.