A causal model is proposed in order to explain secondary students' achievement in chemistry; the seven predictor variables used in the explanation are: students' attitude towards chemistry, understanding of science, scientific attitude, traditional cosmological world-view, cognitive style, teachers' attitude towards the teaching of chemistry and understanding of science. Some 460 students chosen by means of stratified cluster sampling and their 25 teachers make up the student and teacher samples, respectively. The numerical values of all the variables are obtained and their intercorrelations computed. Path analysis procedure is applied to the data to estimate the coefficients of the structural equations of the proposed model. The results of the analysis indicate that the hypothesised model is tenable. The seven predictors directly account for 21 6% of the variance in the criterion; their indirect effect is 8 4%. The implications of the findings for chemistry instruction, teacher education and curriculum development are discussed.