This research aimed to determine the effect of Problem Based Instruction learning model with Predict-Observe-Explain approach on chemistry learning outcomes. The population is XI grader Sciences of Senior High School in Brebes for academic year 2011/2012. Initial data analysis showed that the population are normally distributed and homogeneous, so the sampling technique which used is cluster random sampling. From this sampling, XI Science-5 used as a control class (Problem Based Instruction learning model without Predict-Observe-Explain approach) and XI Science-1 as an experiment class (Problem Based Instruction with Predict Observe Explain approach). Final data analysis showed that learning outcomes for both classes are normally distributed and have equal variances. In the correlation test, obtained 0.433 of r b value, which showed a middle correlation, so Problem Based Instruction with Predict-ObserveExplain approach has middle effect on chemistry learning outcomes in solubility and solubility product. This learning contributes to student learning outcomes is 19%. The average value of affective and psychomotor in experimental class is better than the control class. Based on this research, we can conclude that Problem Based Instruction with Predict-Observe-Explain approach has a positive effect on chemistry learning product in Senior High School students. Key Words: Problem Based Instruction Learning