As societies are ageing and at the same time we are faced with other demographic changes like increased female labour force participation and decreasing fertility rates, the involvement of men and women in unpaid caregiving can be expected to increase the next coming decades. Even though caregiving may be considered as good to society, it may induce considerable opportunity costs in terms of reduced employment and increased mental health problems. Especially, when these opportunity costs are caused by caregiving per se it may have major consequences on the formal care market and for the social security system. Analysing the first two waves of the Survey of Health Ageing and Retirement (SHARE) we examine the effect of informal caregiving on employment and mental health in an Instrumental Variables framework. We also take advantage of the panel data structure to control for time-invariant individual heterogeneity by employing a Fixed Effects model. Our Fixed Effects results suggest a significant negative effect of informal caregiving on the employment probability of 5-6pp for females only. Examining the effect of caregiving on mental health, the Random Effects results suggest a significant positive effect on the total number of depressive symptoms, ranging from 0.20 (CARE) to 0.48 (PARCARE) also for females only. In the male sample the instrumental variables could not be properly used to identify the caregiving decision, suggesting that we possibly require other types of instrumental variables to examine the effect for males.