Two distinct approaches can be employed in the classification and divisions of wet/dry climate. The first is the rule-driven strategy with predefined threshold values representing climate division boundaries, which is usually performed by Aridity Index (AI). The second is an automated method in a data-driven fashion, avoiding the direct specification of classification rules while utilizing some forms of cluster analysis. However, various methods for climate classifications raise issues on their applicability and quality. Therefore, evaluation and comparative studies are needed to handle such issues first, in order to analyze their performance and second, to better understand climate characteristics in a region. This article makes a comprehensive analysis and comparison among five classification methods, including four rule-driven methods based on different categories [Penman-Monteith (PM), Thornthwaite, Holdridge, Sahin's method] and one data-driven method (factor-cluster analysis). With the meteorological data for long-term period (1981–2010), the wet/dry climate divisions were performed for 191 meteorological stations in Northwest China (NW). The results indicated that the overall climate regimes were in agreement for five classifications, but boundaries of wet/dry climate divisions in a data-driven fashion showed a better consistency with topographic features. PM classifications displayed more arid climate types in NW, while the Thornthwaite approach showed an underestimate in arid environments. All wet/dry climate types corresponding to Holdridge and Sahin's classifications were represented in NW, but Holdridge classification is more easily affected by topography and elevation. Complete comparisons among rule-driven methods are difficult to conduct due to different class definitions and low coincident classes. Class definition of climate types for different rule-driven classifications thus needs further investigation. This article highlights the importance of acknowledging the limitations and advantages of different classification systems as well as the dry and wet climate conditions in NW in hope to provide a reference for a similar geographical region.