The examination of children's sleep-related mental experiences presents many significant challenges for researchers investigating the developmental trajectories of human dreaming. In contrast to the well-explored developmental patterns of human sleep, data from dream research are strikingly divergent with highly ambiguous results and conclusions, even though there is plenty of indirect evidence suggesting parallel patterns of development between neural maturation and dreaming. Thus results from studies of children's dreaming are of essential importance not only to enlighten us on the nature and role of dreaming but to also add to our knowledge of consciousness and cognitive and emotional development. This review summarizes research results related to the ontogeny of dreaming: we critically reconsider the field, systematically compare the findings based on different methodologies, and highlight the advantages and disadvantages of methods, arguing in favor of methodological pluralism. Since most contradictory results emerge in connection with descriptive as well as content related characteristics of young children's dreams, we emphasize the importance of carefully selected dream collection methods. In contrast nightmare-related studies yield surprisingly convergent results, thus providing strong basis for inferences about the connections between dreaming and cognitive emotional functioning. Potential directions for dream research are discussed, aiming to explore the as yet unraveled correlations between the maturation of neural organization, sleep architecture and dreaming patterns.