Many novel oncology dose escalation designs have been proposed to find the maximum tolerated dose (MTD), but only a small number of designs based on the dose limiting toxicity (DLT) data have been widely implemented in real clinical trials, including rule-based designs (e.g., i3+3), model-based designs (e.g., CRM and BLRM) and model-assisted designs (e.g., mTPI, mTPI-2, and BOIN). In this chapter, we have a brief review of popular novel designs, discuss the challenges in their applications for phase 1 oncology trials, and provide practical guidance on the design selection, dose escalation decision-making, and the final MTD determination. Specifically, the prior specification for the model-based designs and the considerations on overdosing control are investigated and discussed. We evaluate different designs from the practical perspective of dose escalation decision-making, conduct comparative analysis in terms of operating characteristics in simulation studies, and provide some recommendations to improve the practice. A new approach to improve the MTD determination by DLT weighting is also presented. Finally, some methods beyond DLT-based dose escalation designs are introduced to handle late-onset toxicity, measure toxicity burden, and balance efficacy and toxicity for dose optimization.