An experiment was conducted in the screen house of the Department of Crop, Soil and Pest Management of the Federal University of Technology, Akure, Nigeria during the 2016/2017 dry season. Drought as a constraint to maize production have caused 24 million tonnes yield loss annually on a global scale, thus the development of maize genotypes that can withstand drought stress at vegetative phase. The objectives of the experiment were to (i) determine the magnitude of genetic variation among ten maize population and the extent of association among them (ii) identify the vegetative traits that could be used in discriminating between drought tolerant and drought susceptible maize genotypes, and (iii) identify the best drought tolerant population (s) to be used as parents in hybridization. The ten maize populations were planted in polythene bags in a split plot experiment laid out in Completely Randomized Design (CRD) with three replications having four drought stresses (1, 2, 3 and 4 weeks of moisture stress) and a control. Data were collected on plant height, root length, number of roots per plant, number of leaves per plant, root weight, biomass and shoot weight. Data obtained were subjected to analysis of variance (ANOVA) and significant means were separated using Tukey Honest Significant Difference. Also, Principal Component Analysis (PCA) was employed to identify the most discriminating characters for the maize populations. Results of the analysis of variance revealed significant variation among the populations for all traits except for emergence rate index, emergence index and Plant height. The PCA identified percentage senescence, root length, number of roots per plant and plant height as the four most discriminating traits contributing most to variation for drought tolerance of the maize populations at vegetative phase. From this study, Rank Summation Index revealed DT-Y-Syn / 4 and TZL Comp.1-WC6/white DT STR (Syn) -DFC as the best and worst populations for drought tolerance.