Kinetic theory suggests that the temperature sensitivity of decomposition of soil organic matter should increase with increasing recalcitrance. This ‘temperature–quality hypothesis’ was tested in a laboratory experiment. Microcosms with wheat straw, spruce needle litter and mor humus were initially placed at 5, 15 and 25 °C until the same cumulative amount of CO2 had been respired. Thereafter, microcosms from each single temperature were moved to a final set of incubation temperatures of 5, 15 and 25 °C. Straw decomposed faster than needle litter at 25 and 15 °C, but slower than needle litter at 5 °C, and showed a higher temperature sensitivity (expressed as Q10) than needle litter at low temperatures. When moved to the same temperature, needle litter initially incubated at 5 and 15 °C had significantly higher respiration rates in the final incubation than litters initially placed at 25 °C. Mor humus placed at equal temperatures during the initial and final incubations had higher cumulative respiration during the final incubation than humus experiencing a shift in temperature, both up- and downwards. These results indicate that other factors than substrate quality are needed to fully explain the temperature dependence. In agreement with the hypothesis, Q10 was always higher for the temperature step between 5 and 15 °C than between 15 and 25 °C. Also in agreement with the temperature–quality hypothesis, Q10 significantly increased with increasing degree of decomposition in five out of the six constant temperature treatments with needle litter and mor humus. Q10s for substrates moved between temperatures tended to be higher than for substrates remaining at the initial temperature and an upward shift in temperature increased Q10 more than a downward shift. This study largely supports the temperature–quality hypothesis. However, other factors like acclimation and synthesis of recalcitrant compounds can modify the temperature response.