A commercial drinkable yogurt with and without 4% of added trehalose (as cell protectant) was spray-dried obtaining a powder with low water activity (aw). Total bacterial count in the powder was between 8.48–8.90 log cfu/g. The dried yogurt was stored: (i) at 38 °C and aw = 0.33; (ii) at 38 °C in hermetically sealed flasks (aw = 0.21/0.22); (iii) in a cyclic temperature chamber (10–20 °C) in hermetically sealed flasks (aw = 0.21/0.22). Whole milk was then fermented by adding an inoculum of spray-dried yogurt after storage under these different conditions. The kinetics of acidification showed the presence of a lag time which was strongly dependent on storage conditions. The data was fitted with a logistic type equation from which the lag time was calculated. To evaluate structural differences among samples, Fourier Transform Infrared spectra (FTIR) were recorded. Partial Least Squares (PLS) models enabled a good correlation between lag time of fermentation and FTIR spectra. The lag time for yogurt powder stored at aw about 0.21/0.22 and cyclic temperature 10–20 °C remained approximately constant over the 12 weeks of storage, while all the other conditions resulted in a dramatic increase. The addition of trehalose had a small influence on lag time and, therefore, as a protectant of lactobacilli.