Can hot water freeze faster than cold water (the Mpemba effect)?

Still disputed in the literature as of 2026.

Named for Tanzanian student Erasto Mpemba, who reported in 1969 (with D. G. Osborne) that hot ice-cream mix froze before cold mix, the effect has since drawn dozens of studies proposing mechanisms including evaporative cooling, dissolved-gas content, convection currents, and supercooling point differences, none of which fully explains all reported observations. Burridge and Linden's 2016 Scientific Reports study argued that claimed demonstrations are extremely sensitive to thermometer placement and container geometry, and that under careful controls hot water does not cool faster, casting doubt on whether a robust, generic Mpemba effect exists at all. The debate remains active, with some researchers defending narrower, protocol-specific versions of the effect and others maintaining it is largely an experimental artifact.

An 'obviously true' bit of playground physics has survived 50+ years of careful experiment without anyone agreeing on whether it is even real.

Explanations that circulated

• The effect is caused simply by faster evaporative mass loss from the hot sample, fully accounting for the reduced freezing time in every reported case
• It is purely a measurement artifact and never occurs under any real conditions

References

• Questioning the Mpemba effect: hot water does not cool more quickly than cold, J. I. Katz-critiqued; H. C. Burridge, P. F. Linden, Scientific Reports, 2016
• The Mpemba effect: When can hot water freeze faster than cold?, Monwhea Jeng, American Journal of Physics, 2006