An exploratory study was conducted in the rural Estie district of Ethiopia in 1997 to identify the role of ABO blood group, rhesus factor, and type of grass pea (Lathyrus sativus) diet in the susceptibility to neurolathyrism. Five-hundred study subjects (250 cases and 250 controls) were examined and interviewed, and had their ABO and rhesus blood groups determined. The majority (86%) of the cases were males. Blood group O was the most common in the patients and controls followed by groups A, B, and AB. The vast majority of the study subjects were rhesus-positive. The gravy (Shiro) grass pea preparation was consumed by 91.6% of the study population, boiled (Nifiro) by 86%, and roasted (Kollo) by 56.4%. Almost half (48%) of the cases had consumed grass pea for > 4 months compared to 8% of controls (P < 0.001). There was a significant association between the risk for neurolathyrism and the consumption of boiled (adjusted odds ratio [AOR] = 98.4) and roasted (AOR = 55.62) forms of grass pea. There was no risk of paralysis associated with consumption of the gravy form of grass pea (AOR = 0.40, 95% confidence interval 0.1-2.0). Blood group O remained significantly associated with the disease after adjusting for age, type of grass pea preparation consumed, and duration of consumption (AOR = 2.90).