To examine the effects of general and specific comprehension ability of a target reading passage and prior knowledge of the subject of the passage, a study examined 182 sixth graders from two central Illinois rural and urban communities. Subjects were given a fifth-grade passage (a 500-word fictional narrative description of the Yanomamo tribe of the Amazonian Basin) from the "De Santi Cloze Reading Inventory." Subjects were given either the original passage, or a difficult version created by substituting a difficult synonym (about eighth-grade level) for every sixth content word. Cloze condition students were given the passage in a conventional cloze format. Other students were given an intact version of the passage. Students were first given a cloze passage from the "De Santi Cloze Reading Inventory" to equate ability level. Before reading, students were told either key facts about the Yanomano villagers, or facts which were tangential to the narrative. Following the presentation students read the passage. In the cloze condition, students filled in the cloze blanks with a word which made sense in the passage. In both conditions, the passage was taken away after reading and subjects were given a multiple choice test assessing the knowledge of 10 of the substituted words. Results confirmed earlier studies indicating that children learn word meanings from context, but did not find effects on the rate of that learning due to task manipulations, prior knowledge, and general or specific comprehension. (One figure, the multiple choice test, and 22 references are appended.) (MM)