Our empirical study analyzes house search duration in ìhotî (rising prices) and ìcoldî (declining prices) housing markets. It focuses on the effect of householdsí ex ante holding horizon ñ or the expected housing tenure at the time of their search ñ on the realized search duration. Our empirical analysis controls for physical and locational housing characteristics, transaction price, financing choices across buyers, buyer preferences, and other factors, and the results are analyzed for a pooled sample, and then compared for the ìhotî and ìcoldî housing markets. To our knowledge, this is the first empirical study of this kind. Large amounts of information available in realtorsí databases, MLS and other real estate data sources, have enabled researchers to analyze in depth various aspects of housing liquidity on the seller side. However, the buyer side analyses are very limited, as this information is often only available through a costly and likely a time-consuming survey. Two surveys of recent homebuyers were performed in the ìhotî market in 2005 and in the ìcoldî market in 2008. In each case, over 6000 questionnaires were mailed in a large geographical area in California. A12% response rate provided enough data for our study. Our analysis shows that adding the expected housing tenure variables improves the overall fit of our empirical model of effects on buyer side liquidity, or time until purchase. In the pooled sample, a longer expected occupancy increases the search time. Furthermore, investors, who occupy about 10% of our sample (13% in ìhotî, 7% in ìcoldî), spent statistically significantly more time buying a house if they plan to sell it in the next 1-5 years, rather than those with longer holding horizons. This result suggests that the effects of expected housing tenure on search duration is non-homogeneous across buyer types as it is sensitive to consumption versus investment purchase intent. An interesting empirical result was obtained when we controlled for the proximity of a house to the coast. In both ìhotî and ìcoldî subsamples households searched longer when the purchased house was located in a coastal zip code area, and this effect was statistically significant. When each subsample was decomposed into the coastal and the inland regions, the investment motive appears to be present only for inland houses while purchasing a coastal property appears to be consumption-driven. Second, the positive tenure dependence effect is highly statistically significant only for coastal houses in the ìhotî market, and inland houses in the ìcoldî market. Does this imply that in the ìhotî market households tend to improve their living standards by ìmoving upî (and thus select their better-located home more carefully if expected housing tenure is longer), while in the ìcoldî market they instead tend to ìmove downî and thus buy houses with careful tenure considerations in more inland regions? We believe that our study would contribute to the literature of the housing liquidity on buyer side, as well as enrich the understanding of different driving forces of house search under different housing market conditions.