The purpose of the present study was to understand the effect of formulation variables of self- nanoemulsified drug delivery systems (SNEDDS) on the rapid dissolution of a model drug, genistein (GN). A three-factor, three-level Box-Behnken design was used to explore the main and interaction effect of several independent formulation variables including the amount of Maisine 35-1 and Labrafac Lipophile WL 1349 (1:1, w/w) (X1), Cremophor EL and Labrasol (3:1, w/w) (X2), and Transcutol P (X3). Droplet size (Y1), turbidity (Y2), and dissolution percentage of GN after 5 (Y3) and 30 (Y4) min were the dependent variables. A mathematical relationship, Y3 = - 89.3447 + 5.9524X1 + 1.0683X2 + 0.462X3 - 0.0825X(1)(2) - 0.0075X(2)(2) - 0.0009X(3)(2) + 0.0104X1X2 - 0.0113X1X3 + 0.0009X2X3 (r2 = 0.9604), was obtained to explain the effect of all factors and their co-linearities on the dissolution of GN at 5 min. Formulation optimization was then performed to maximize dissolution percentage of GN at 5 min (Y3). The optimized formulation was predicted to dissolution 93.34% of GN at 5 min, when X1, X2 and X3 values were 37.1, 101.7 and 77.3 mg, respectively. A new batch was prepared according to the optimized formulation, and the observed and predicted values of Y3 were in close agreement. In conclusion, the Box-Behnken experimental design allowed us to understand the effect of formulation variables on the rapid dissolution of GN from SNEDDS, and optimize the formulation to obtain a rapid drug dissolution at 5 min.