Longhole stoping is a productive exploitation technique for underground metalliferous deposits. The method consists of deep-hole blasting with increased explosive charges to excavate larger volume of ore from a single blast. However, large scale blasting produces hazards in terms of blast-induced vibration. Theunderground structures viz. drivages, cross-cuts, shaft pillar, shaft, decline, etc. in the proximity of blasting face need to be safeguarded from damages due to blast vibration. This paper is a case study on investigating the influence of the charging parameters on the safety of nearby underground structures at Sindesar-Khurd Lead- Zinc underground mine. Experimental blasts with variations in charging parameters were performed at the mine for this purpose. The ground vibration data were recorded from the experiment and analysed using neural network analysis and regression analysis. The importance of different charging parameters on ground vibration magnitude has been identified using importance analysis. Regression analysis has been carried out to establishpredictor equations correlating different charging parameters with peak particle velocity (PPV). The influence of charging parameters has also been investigated using Probability-value (P-value) approach. The neural network and regression analysis approaches have identified the zones from the blasting face where, distance, maximum charge weight per delay, total explosive charge in a ring and total explosive charge in a blasting round has significant impact. Accordingly, the predictors consisting of different charging parameters have been established.The optimum charging parameters have been suggested based on these predictors for blasting at the mine. The suggested charging parameters are focused on to reduce vibration near underground structures within thesafe limit.