Background: Malaria is seasonal and this may influence the number of children being treated as outpatients in hospitals. The objective of this study was to investigate the degree of seasonality in malaria in lakeshore and highland areas of Zomba district Malawi, and influence of climatic factors on incidence of malaria.

Methods: Secondary data on malaria surveillance numbers and dates of treatment of children <5 years of age (n = 374,246) were extracted from the Zomba health information system for the period 2012–2016, while data on climatic variables from 2012 to 2015 were obtained from meteorological department. STATA version 13 was used to analyse data using non-linear time series correlation test to suggest a predictor model of malaria epidemic over explanatory variable (rainfall, temperature and humidity).

Results: Malaria cases of children <5 years of age in Zomba district accounts for 45% of general morbidity. There was no difference in seasonality of malaria in highland compared to lakeshore in Zomba district. This study also found that an increase in average temperature and relative humidity was associated of malaria incidence in children <5 year of age in Zomba district. On the other hand, the difference of maximum and minimum temperature (diurnal temperature range), had a strong negative association (correlation coefficients of R2 = 0.563 [All Zomba] β = −1295.57 95% CI −1683.38 to −907.75 p value <0.001, R2 = 0.395 [Zomba Highlands] β = −137.74 95% CI −195.00 to −80.47 p value <0.001 and R2 = 0.470 [Zomba Lakeshores] β = −263.05 95% CI −357.47 to −168.63 p value <0.001) with malaria incidence of children <5 year in Zomba district, Malawi.

Conclusion: The diminishing of malaria seasonality, regardless of strong rainfall seasonality, and marginal drop of malaria incidence in Zomba can be explained by weather variation. Implementation of seasonal chemoprevention of malaria in Zomba could be questionable due to reduced seasonality of malaria. The lower diurnal temperature range contributed to high malaria incidence and this must be further investigated.