The authors discuss the differences between the two large series of chronic childhood spinal muscular atrophies (SMA)--their own comprising 273 cases, and that of Pearn et al. comprising 141 cases. The main difference concerns the predominance of males in the clinically milder later-onset group in the present series. The data of Pearn et al. (1978a, b) are quite different. The reason for the discrepancies is apparently a different selection of material. The present material is highly selective in favour of chronic cases, and Kugelberg-Welander cases are well-represented, whereas the percentage of Kugelberg-Welander cases in the material of Pearn et al. was very small. Differences in selection also appear to be responsible for discordance in observations regarding influence of sex on the course of the disease. The present data seem to support the view that most of the cases revealing chronic forms of SMA (both mild and severe) are not distinct genetically. However, the possible existence of a distinct subgroup in which sex influence is strongly expressed is not excluded.