A new mechanistic deterministic model for assessing the impact of temperature variability on malaria transmission dynamics is developed. Sensitivity and uncertainty analyses of the model parameters reveal that, for temperature values in the range 16-34 degrees C, the three parameters with the greatest influence on disease dynamics are the mosquito carrying capacity, transmission probability per contact for susceptible mosquitoes and human recruitment rate. This study emphasizes the combined use of mosquito-reduction strategies and personal protection against mosquito bites during periods when the mean monthly temperatures are in the range 16.7-25 degrees C. For higher monthly mean temperatures in the range 26-34 degrees C, mosquito-reduction strategies should be emphasized ahead of personal protection. Numerical simulations of the model reveal that mosquito maturation rate has a minimum sensitivity (to the associated reproduction threshold of the model) at 24 degrees C and maximum at 30 degrees C. The mosquito biting rate has maximum sensitivity at 26 degrees C, while the minimum value for the transmission probability per bite for susceptible mosquitoes occurs at 24 degrees C. Furthermore, it is shown, using mean monthly temperature data from 67 cities across the four regions of sub-Saharan Africa, that malaria burden (measured in terms of the total number of new cases of infection) increases with increasing temperature in the range 16-28 degrees C and decreases for temperature values above 28 degrees C in West Africa, 27 degrees C in Central Africa, 26 degrees C in East Africa and 25 degrees C in South Africa. These findings, which support and complement a recent study by other authors, reinforce the potential importance of temperature and temperature variability on future malaria transmission trends. Further simulations show that mechanistic malaria transmission models that do not incorporate temperature variability may under-estimate disease burden for temperature values in the range 23-27 degrees C, and over-estimate disease burden for temperature values in the ranges 16-22 degrees C and 28-32 degrees C. Additionally, models that do not explicitly incorporate the dynamics of immature mosquitoes may under- or over-estimate malaria burden, depending on mosquito abundance and mean monthly temperature profile in the community.