Research into boiling has been carried out over the last several decades and extensive data have accumulated from experimental studies carried out under various conditions and configurations, leading to the development of the currently available empirical and phenomenological correlations. However, most correlations apply to situations under a relatively narrow range of conditions and exhibit a considerable error band, even for the data sets on which they are based. In contrast to this multitude of correlations, the development of mechanistic models based on the underlying physical processes has been sporadic and limited. Thus, it is said that the use of boiling is not yet a "mature" technology, though it is one of our important technologies. This paper summarizes recent developments in research addressing the nonlinear dynamical behaviors of pool nucleate boiling for the purpose of suggesting some potential reasons why we have had limited success in mechanistic modeling, and to provide a promising perspective on future boiling research. In the introductory section, the problems facing current boiling research are described. In the following two sections, recent nonlinear experimental as well as theoretical research achievements are overviewed, while in the last section, the problems that remain unsolved are discussed, along with some concluding remarks. (C) 2003 Elsevier Ltd. All rights reserved.