Because figural reasoning tasks are often assumed to indicate fluid intelligence (gf), we investigated which aspect of figural reasoning tasks make that they tend to mark gf-the figural content itself or the high degree of abstraction in these tasks. To this end, the assessment of figural reasoning abilities by means of concrete figural task material was compared with the assessment of figural reasoning abilities using abstract figural task material, and the relations of the abstract and the concrete figural reasoning tasks to fluid (gf) and crystallized intelligence (gc) were investigated. The central hypothesis was that figural reasoning tasks containing semantically meaningful objects would show higher gc-loadings than abstract figural reasoning tasks. Three newly developed figural reasoning tasks with semantically meaningful material were administered to 144 German subjects, along with the German intelligence test I-S-T 2000 R (Amthauer, R., Brocke, B., Liepmann, D., & Beauducel, A. (2001). Intelligenz-Struktur-Test 2000 R. Gottingen: Hogrefe) containing abstract figural reasoning tasks as markers for gf and knowledge items as markers for gc. Confirmatory factor analysis revealed that concrete figural reasoning tasks exhibited larger gc-loadings than abstract figural reasoning tasks. (C) 2004 Elsevier Inc. All rights reserved.