An objective multivariate technique is described for identification of individual tigers Panthera tigris from their pugmarks. Tracings and photographs of hind pugmarks were obtained from 23 pugmark-sets of 19 individually known tigers (17 wild and two captive tigers). These 23 pugmark-sets were then divided into two groups, one of 15 pugmark-sets for model building and another of eight pugmark-sets for model testing and validation. A total of 93 measurements were taken from each pugmark along with three gait measurements. We used CV ratio, F-ratio and removed highly correlated variables to finally select 11 variables from these 93 variables. These 11 variables did not differ between left and right pugmarks. Stepwise discriminant function analysis (DFA) based on these 11 variables correctly classified pugmark-sets to individual tigers. A realistic population estimation exercise was simulated using the validation dataset. The algorithms developed here correctly allocated each pugmark-set to the correct individual tiger. The effect of extraneous factors, i.e. soil depth and multiple tracers, was also tested and pugmark tracings compared with pugmark photographs. We recommend collecting pugmarks from soil depths ranging between 0.5 and 1.0 cm, and advocate the use of pugmark photographs rather than pugmark tracings to eliminate the chance of obtaining substandard data from untrained tracers. Our study suggests that tigers can be individually identified from their pugmarks with a high level of accuracy and pugmark-sets could be used for population estimation of tigers within a statistically designed mark–recapture framework.