Chemistry learning in high school should encourage the application of knowledge learned to solve real-world problems around students without delay. This study aimed at analyzing the effect of the project based e-learning model (PjBeL) compared to direct e-learning (DeL) and academic procrastination (AP) on students' critical thinking skills and cognitive achievement. This quasiexperimental research used the pretest-posttest nonequivalent control group design. The study population was 5 classes (148 students) XI MIPA of SMAN 1 Bebandem. The sample was selected using the class rendom technique to determine 4 classes (120 students, or 81.1% of the population) as reseach samples. This technique is also used to determine 2 classes as the experimental group and the other 2 classes as the control group. Academic procrastination data were collected by using questionnaires, critical thinking skills data with essay tests, while learning achievement data with expanded multiple choice tests. The data were analyzed descriptively and multivariate analysis of covariance. The results showed that, (1) there were differences in critical thinking skills and student achievement between students who studied with those models, in which the students who studied with the PjBeL model showed a better result. (2) The differences of students' AP had an impact on students' achievement, but did not affect the achievement of their critical thinking skills. students with low academic procrastination (LAP) achieve higher learning achievement than those with high academic procrastination (HAP). (3) There is no interactive effect between the learning model (PjBeL v.s DeL) and academic procrastination (LAP v.s HAP) on both student achievement and critical thinking. The implication of this research is that in achieving students' achievement and their critical thinking, chemistry learning will be better using PjBeL.