A study of hospital admissions of chronic renal failure cases during the 5 years 1987-1991 in a major referral centre in South West African showed a significant departure from a uniform monthly distribution (P < 0.01). Admission rates were high in February-March and June-August showing 2 peaks. There was no absolute seasonal effect. Mean monthly admissions did not differ between the dry (9.7 +/- 0.6 SE) and wet (9.8 +/- 0.6 SE) seasons. Possible reasons for the variations are discussed. For health planning, high admission rates can be anticipated and resources mobilized to meet the increased demands at these periods when they occur.