The household …nance literature documents a large fraction of the population not participating in stock markets. It is also puzzling that a much greater share of households do not participate in foreign stock markets. Recent empirical evidence points towards the role of information in determining agents'portfolio choices. I test these results into a model that incorporates information on agents' portfolio allocation decision. Departing from Abel at al (2007), consumers can invest in foreign stocks, in addition to domestic ones. To update their information set, agents have to pay a cost implying that consumers update their portfolio only infrequently. In addition, to account for the initial costs of acquiring information about stock investments, a version of the model also features an entry-cost to be paid at the …rst period by agents that decide to enter stock market. Agents that invest in foreign stocks are more attentive, updating their portfolio more frequently. After calibrating the model to match returns and volatility for the U.S. economy and di¤erent foreign stock investments, I obtain that the minimum entry cost necessary to drive households completely out of stock markets is large (and in line with the equity premium puzzle literature). However, once agents already invest in domestic stock markets, the minimum cost that would drive investors out of foreign stocks market is much smaller. The size of the latter minimum entry cost depends on model parameters assumptions, and small variations on risk aversion and uncertainty about foreign asset returns can make this entry cost small enough to rationally justify the substantial non-participation in foreign stock markets. JEL classi…cation codes: G11, G14, G15