We study the impact of work from home (WFH) on subjective well‐being during the Covid period, where self‐selection of individuals into telework is ruled out, at least part of the time, by stay‐at‐home orders. We use a difference‐in‐differences approach with individual fixed effects and identify the specific impact of switching to telecommuting, separately from any other confounding factor. In particular, our identification strategy avoids the influence of interpersonal heterogeneity by exploiting the multiple entries into WFH, by the same individuals, at different times. On average over the period, switching to WFH, especially full‐time, worsens mental health. We also find a positive but imprecisely measured impact of part‐time WFH on life satisfaction. However, this hides a dynamic evolution, whereby the initial deterioration gives place to an adaptation process after a couple of months. We also uncover a particularly pronounced fall in subjective well‐being of women with children, especially in the first months; this could be associated with home‐schooling.