Mixed dentition space analysis is very important in orthodontic diagnosis and treatment planning. The present study evaluated prediction methods for the estimation of size of unerupted canine & premolars, and observed any sexual dimorphism in tooth size in a local population. A cross-sectional study was conducted on 192 boys and 192 girls aged 13-15 years who met the inclusion /exclusion criteria from schools of Lahore, Pakistan. Dental impressions of both arches were made and mesiodistal widths of teeth were measured with digital vernier caliper. Students t-test, Pearson correlation coefficient analysis and simple linear regression analysis was used to statistically compare tooth size differences in arches, among genders and predict maxillary and mandibular canine and premolars width based on sum of four mandibular incisors; sum of mandibular incisors and first molar; and sum of mandibular incisors and maxillary first molar. New regression equation was formulated for each arch and sex separately based on sum of mandibular incisors and maxillary first molar. Sum of maxillary first molar and mandibular incisors showed high correlation and determination coefficient in prediction of size of unerupted canine and premolars in the study population. Significant gender difference was found in tooth size in this sample. Sum of mandibular incisors and maxillary first molar is the better predictor for estimation of size of unerupted canine and premolars in our study population. Proposed prediction equations showed good accuracy, easy application and can be used in orthodontic treatment planning in local population.