Mpemba effect: does hot water freeze faster than cold water

Still disputed in the literature as of 2026.

Aristotle noted in antiquity that hot water can cool and freeze faster than cold water, but the phenomenon became widely known as the Mpemba effect after Tanzanian student Erasto Mpemba observed it making ice cream in 1963 and, with physicist Denis Osborne, published on it in 1969. Since then dozens of studies have proposed mechanisms (evaporative mass loss, differential supercooling, dissolved gas content, convection) with conflicting results. In 2016 Henry Burridge and Paul Linden at Cambridge ran careful, well-controlled experiments and found no evidence of a genuine, repeatable effect, arguing much earlier literature suffered from poorly controlled protocols, leaving the field in active, unresolved dispute.

Whether hot water can genuinely out-freeze cold water is a question ancient enough for Aristotle to have written about, yet modern physicists with electronic thermometers still cannot fully agree whether the effect is even real.

Explanations that circulated

• It is often claimed the effect is simply explained by evaporation, supercooling avoidance, dissolved gas content, or convective mixing; these have all been proposed as partial contributors, but none is a settled, generally-accepted quantitative explanation, and Burridge & Linden's controlled experiments found no meaningful, repeatable effect at all under carefully controlled conditions.
• Some popular science accounts present the Mpemba effect as a firmly established, mainstream-accepted phenomenon with a known cause; there is no scientific consensus that it reliably occurs under well-controlled conditions, let alone consensus on why.

References

• Questioning the Mpemba effect: hot water does not cool more quickly than cold, Henry C. Burridge, Paul F. Linden, Scientific Reports, 2016