Children who come from socio-economically advantaged homes usually do well in school and vice versa. Further, we would expect that good pupils in the 3rd grade of elementary school (good in a sense of a high academic achievement) are good readers and good readers are expected to be high achievers at school as well.At PIRLS 2001 (Progress in International Reading Literacy Study 2001) we have found that in Slovenian elementary schools things are not exactly like expected. There were many students whose performance at PIRLS was (significantly) above the average achievement but their school grades were 3, 4 or lower (5 is the best grade). On the other hand we have found many students whose reading achievements were beneath the average achievement but their school grades were excellent. According to their school success and according to the reading achievement at PIRLS we have sorted children into four groups: (1) good readers and good students, (2) poor readers and poor students, (3) good readers and poor students and (4) poor readers and good students. Groups (1) and (2) seems not to be problematic. Groups (3) and (4) demand further research. Group (3) is problematic due to the following reasons: Good readers understand what they read, however, they are not able to 'give back' what is demanded at school and even if writing skills had not been checked out at PIRLS students had to show at least basic writing communication skills at open ended questions. Obviously the school rewards not understanding but something else. But what is school rewarding? There are a lot of things: cooperation, good i.e. not conflict behavior, reproduction of what teacher have said (good memory), etc. Group (4) can also be interpreted in many ways: students had bad day at the day of the PIRLS testing or as said for group (3): teacher might reward 'saying without understanding' (we can not tell it is knowledge).In this paper I would like to put the latter question upside-down and examine the group (3): Why is school not rewarding understanding at 3rd graders (their average age at the time of testing was 9.8 years)? I will present background of children who are not doing well at school but are good readers in a sense that they understand what they read without help of an adult (e.g. teacher). What are characteristics of these children: gender, family income, immigration status, parent's occupation, parent's interest in school, cooperation between school and home, teachers' education, children' self-esteem. I will present what are possible influences of the background characteristics on the school grades. The question which is sought to be answered is: are there any arguments to assert the possibility that prejudices have something to do with school grading or can we deny that and only conclude that bad grades are only circumstantial and random occurrences that happen because of wrong expectations (of what school is rewarding)?_________________________