Landslide hazard assessment portraying spatial distribution has gained widespread recognition over the past two decades. This paper presents two methods of landslide hazard evaluation namely; Analytic Network Process (ANP) and Competitive Neural Networks (CNN) techniques. The ANP is a qualitative approach founded on the subjective opinion of experts, whereas, the CNN is a knowledge-based unsupervised learning algorithm of Artificial Neural Networks family. Using the methods, landslide hazard maps were produced of an area in a fragile mountainous terrain in the eastern part of Nepal Himalayas. The results from two methods were compared and validated with the spatial slope failure data. Although the hazard maps produced from both methods are comparable, and closely indicative of the factual landsliding in the field, the ANP method is more convincing because it utilizes subjective judgments of geotechnical experts and establishes interdependence among variables.