Developed was a "reading free" test to evaluate knowledge and understanding of science by elementary school pupils. This study was made because it was suspected that students reading below grade level were unable to indicate fully their science knowledge on conventional tests. The Pictorial-aural Inventory of Science Knowledge (PISK) was designed with content, vocabulary and test items similar to the Reading Inventory of Science Knowledge (RISK). Three hundred students (equal numbers of boys and girls) were divided into groups--average, below average, and above average as indicated by standardized reading test scores. The administration of PISK and RISK were alternated so that the practice effect of one test upon another would be minimized. Conclusions drawn were (1) PISK was found to be a reading free instrument useful in assessing student's understanding and knowledge of elementary school science, (2) RISK scores were highly correlated with reading ability, (3) when reading was controlled, the below average students scored almost as high as the above average group, and (4) boys scored higher on PISK than girls, but not on RISK. (BC)