The aim of this study is to find basic quantitative geometric data that may contribute to the understanding of the etiology of hallux valgus. Embalmed specimens with existing hallux valgus (N = 39) were dissected; 28 variables were measured with a Vernier caliper gauge and toe goniometer. Correlations between pairs of independent variables were calculated. Linear dependency of the hallux angle, varus angle, and the width of the forefoot on a number of independent variables was analyzed by multiple linear regression. A least squares method and a stepwise procedure were used. The distance from the tendon of the flexor hallucis longus muscle to the head of the first metatarsal bone explains more than other variables the variation in hallux angle and width of the forefoot. A widened forefoot is significantly correlated with both hallux and varus angles. The interrelation of the predictor parameters illustrates the complicated hallux valgus phenomenon.