The study aimed to determine the most effective of three pedagogies to improve Grade 12 learners' understanding of the photoelectric effect when the teaching was done using chalk and talk, guided learning with computer simulations (CS) as a demonstration tool and semi-autonomous learning where the learners used the CS individually with limited support. A randomised pre-/post-test control group design was implemented. A paired-sample t-test on the post-test scores indicated that the guided learning experimental group (N = 10) performed better than the semi-autonomous learning experimental group (N = 10) [t-statistics, t(9) = -6.135, p < 0.05]. Even the control group (N = 10) where the teacher used chalk and talk performed better than the semi-autonomous learning experimental group. An analysis of covariance on the post-test scores with learners' pre-test scores showed a significant effect on the instructional group favouring the guided learning experimental group. It is, therefore, suggested that guided learning where the teacher manipulates the CS is a better method for the effective teaching of the photoelectric effect.