A non‐Jewish child with late onset GM2gangliosidosis is described. Tissues from the patient had near normal hexosaminidase A (hex A) activity using 4‐methylumbelliferyl‐2‐acetamido‐2‐deoxy‐β‐D‐glucopyranoside (MU‐glcNAc) as substrate, and deficient activity when assayed with the 6‐sulfate derivative of MU‐glcNAc (MU‐glcNAcS) or GM2in the presence of activator. We present evidence that this patient is a genetic compound for different oc‐subunit mutations. The father's tissues have hex A activity in the heterozygote range when assayed with MU‐glcNAcS, but normal activity using MU‐glcNAc; the mother's tissues have activities toward both substrates in the heterozygote range. These results emphasize the pitfalls of using only MU‐glcNAc for the diagnosis of unu