: The aims of this research are to test the effectiveness of CIRC learning model and the reading ability of the students writing narrative skill. The method that is use in this reseacrh was kuasi experiment and the research design was pre-test, post-test, non-equivalent control group design (Factorial 2x2). The research population was the seventh grade students of SMP N 1 Kasreman and the seventh grade students of SMP N 1 Padas in the academic year 2011/2012. The sample consist of two groups, they were the students of class VII-B SMP N 1 Kasreman and the students of class VII-C SMP N 1 Padas. As control groups is the students of class VII-C SMP N 1 Kasreman and the students of class VII D SMP N 1 Padas as experiment groups. The sample was took by using cluster random sampling. Dependent variabel was writing narrative skill, while independent variables were learning model (CIRC learning model and conventionla learning) and reading ability (high and low). The results of this research are: (1) there was a significant difference of the influence between the implementation of CIRC learning model and convensional learning model in students writing narrative skill that shown byanalisis result T-test (th> tt or 3,423>1,980); (2) there was not significant differenceopf the influence between high reading ability and low reading ability in students writing narrative skill that shown by analysis result T-test (th> tt or 13,321>1,980); and (3) there was not influence interaction between the implementation of learning model and the reading ability in students writing narrative skill that shown by 2 path ANOVA (F0>Ft or 103,41> 3,94; F0>Ft or 80,17 > 3,94; F0>Ft or 96,53> 3,94; F0>Ft or 97,57> 3,94).