Earthquake-induced landslides and liquefaction are important secondary earthquake perils that can cause substantial damage to the built environment in addition to direct damage caused by seismic ground shaking. In spite of their impacts, they are not regularly included in probabilistic seismic hazard and risk analysis (PSHRA), in part because they have not been incorporated in most PSHRA frameworks such as GEM’s OpenQuake Engine. As part of the TREQ project, existing landslide and liquefaction models were implemented within the OpenQuake Engine, and have been made available for both probabilistic and deterministic (scenario) analyses. In this study we present the methodological approach we used to implement these models using the city of Cali as the case study. Regarding coseismic landslides, found that the probability of coseismic landslides within the city limits of Cali is extremely small, although it is likely higher in the adjacent mountain regions. For liquefaction analysis, we tested the models on seismic scenarios selected by the USGS through a hazard disaggregation process. The risk metrics obtained suggest that, in the case of liquefaction, the models make an appropriate prediction of the spatial distribution of damage and loss. However,in terms of the absolute number of damaged structures, estimates for both, landslide and liquefaction risk, are inconsistent with the level of damage and loss obtained from the ground shaking. Hence, we concluded that the existing methodologies do not perform satisfactorily in urban risk applications.