The generative learning strategy of learning by drawing has received increased attention in recent years. Although this strategy is regularly used by educators, the literature suggests that the effectiveness of the method depends on several factors. In this review, we highlight recent research trends and methodological progress within the field. Although recent developments have led to clearer and more comparable results, many current studies still revealed contradicting results in terms of the efficacy and boundary conditions of the learning strategy. In terms of study designs and the targeted types of knowledge, a trend towards digital drawing studies and a growing variety regarding the content domains of the investigated learning tasks can be observed. Based on the fact that results between different studies are often difficult to compare, we argue to more clearly differentiate visualizations and standardize the terminology for visualizations utilized in learning-oriented drawing tasks. As the properties of visualizations can cause varying cognitive demands on learners, differences in the level of skill required to produce different types of drawings, among other factors and learner variables, can affect the outcomes of this teaching method. Based on our review, we discuss practical and ethical implications as well as considerations for future research.