Three incubation experiments were carried out with a non-saline soil (electrical conductivity in a saturation paste (ECe) 1 dS m−1) to which NaCl was added to achieve ECe 10 and 30 dS m−1; pea straw was added at 20 g kg−1 as a nutrient source. Experiment 1 showed that cumulative respiration was highest in soil EC 1 and lowest in soil EC 30. The optimal water content for respiration was 60–70 % of WHC in all soils. There were two periods (days 1–7 and days 8–17) in Experiment 2. In the treatments with the same water content in both periods [optimal (O-O) and medium (M-M)], respiration rates decreased over time and were lower in M-M than in O-O. Cumulative respiration at medium water content did not differ between slow (L-SM) or rapid rewetting (L-RM) from low to medium water content. There were two periods in Experiment 3 with the water content in the first period 50, 40 or 30 % of WHC adjusted from 60 % during pre-incubation either slowly or rapidly. The water content in the second period was maintained or adjusted slowly to 30–60 %. Cumulative respiration differed between water contents but was not consistently different between rapid and slow drying in the first period. We conclude that the response of microbial activity to a certain water content is influenced by the previous water content whereas the speed at which the water content is adjusted had little effect on respiration at target water content.