This is a third exploration of risk factors for the two most severe forms of fetal alcohol spectrum disorders (FASD), fetal alcohol syndrome (FAS) and Partial FAS (PFAS), in a South African community with the highest reported prevalence of FAS in the world. In a case control design, interview and collateral data concerning mothers of 72 first grade children with FAS or PFAS are compared with 134 randomly selected maternal controls of children from the same schools. Significant differences were found between the mothers of FASD children and controls in socio-economic status, educational attainment, and a higher prevalence of FASD among rural residents. The birth order of the index children, gravidity, and still birth were significantly higher among mothers of FASD children. Mothers of children with a FASD are less likely to be married and more likely to have a male partner who drank during the index pregnancy. Current and gestational alcohol use by mothers of FASD children is bingeing on weekends, with no reduction in drinking reported in any trimester in 75 to 90% of the pregnancies that resulted in an FAS child or during 50 to 87% of PFAS-producing pregnancies. There was significantly less drinking among the controls in the second and third trimesters (11 to 14%). Estimated peak blood alcohol concentrations (BAC)s of the mothers of PFAS children range from 0.155 in the first trimester to 0.102 in the third, and for mothers of FAS children the range is from 0.197 to 0.200 to 0.191 in the first, second, and third. Smoking percentage during pregnancy was significantly higher for mothers of FASD children (82 to 84%) than controls (35%); but average quantity smoked is low in the 3 groups at 30 to 41 cigarettes per week. A relatively young average age of the mother at the time of FAS and PFAS births (28.8 and 24.8 years respectively) is not explained by early onset of regular drinking (mean = 20.3 to 20.5 years of age). But the mean years of alcohol consumption is different between groups, 16.3, 10.7, and 12.1 years respectively for mothers of FAS, FASD, and drinking controls. Mothers of FAS and PFAS children were significantly smaller in height and weight than controls at time of interview. The child's total dysmorphology score correlates significantly with mother's weight (-0.46) and BMI (-0.39). Bivariate correlations are significant between the child's dysmorphology and known independent demographic and behavioral maternal risk factors for FASD: higher gravidity and parity; lower education and income; rural residence; drinks consumed daily, weekly, and bingeing during pregnancy; drinking in all trimesters; partner's alcohol consumption during pregnancy; and use of tobacco during pregnancy. Similar significant correlations were also found for most of the above independent maternal risk variables and the child's verbal IQ, non-verbal IQ and behavioral problems. Maternal data in this population are generally consistent with a spectrum of effects exhibited in the children. Variation within the spectrum links greater alcohol doses with a greater severity of effects among children of older and smaller mothers of lower socio economic status in their later pregnancies. Prevention is needed to address known maternal risk factors for FASD in this population.