The influence of various covariates (including weight, race, smoking, gender, age, mild-to-moderate alcohol intake, and body surface area) on the population pharmacokinetic parameters of phenytoin in adult epileptic patients in South Africa was investigated. The parameters were the maximum metabolic rate (Vm) and the Michaelis-Menten (MM) constant (Km) of phenytoin. The study population comprised 332 black and colored epileptic patients (note: "black" refers to indigenous people of South Africa, who speak one of the Bantu languages as their native language; "colored" refers to people considered to be of mixed race, classified as such by the apartheid former government of South Africa). The influence of covariates on Vm and Km estimates was determined using nonlinear mixed-effects modeling (NONMEM). Parameter models describing the factors that could potentially influence Vm and Km were tested using the Michaelis-Menten parallel MM and first-order elimination models, to which 853 steady state dose-to-serum concentration pairs were fitted. The results indicated that body weight, smoking, race, and age (65 years or older), in descending order of importance, significantly influenced Vm (p< 0.05). Although a significant difference (p= 0.03) in Km was found between black and colored patients, incorporating the influence of race in Km in the final regression model did not improve the fit of the model to the data, which indicated that the variability in Km was accounted for by Vm. The scaling factors for smoking, colored patients and age (65 years or older) in Vm were 1.16, 1.10, and 0.88, respectively. These factors should be taken into account when adjusting phenytoin dose.