The full-fledged Japanese census of commerce was conducted in 2014 and its data were publisized recently. We have chased the census data since 1991 in order to explore the determinants of Japanese household expenditure on consumer goods specialty retailers. The purpose of this study is to add some new findings to our previous research. In this research we theoretically address, and empirically estimate, key factors that affect sales per household at three major lines of retail trade that include frequently purchased consumables (food and drink), less frequently bought non-durables (apparel, shoes and dry goods), and infrequently acquired durable goods (furniture). We examine Industrial Classifications 57-60: Dry Goods, Apparel and Apparel Accessory stores (largely clothing, shoe, linen and accessories); Food and Beverage stores (primarily grocery, liquor, and specialty food stores); and Furniture, Household Utensils, and Appliances. These three trade lines deal with the necessities and supplies of life; they consist of relatively small specialty retailers. In addition, they have been an important target of urban planning and retail distribution policy of cities in Japan. Our data, which is drawn from six successive Japanese retail trade censuses (1991, 1994, 1997, 2002, 2007, 2014) encompasses 790 cities in all 47 prefectures. It is notable that the Japanese babble economy ceased in 1991; since then it has experienced an extended period of low growth. Note also that data from the most recent census (2014) is not yet available. Our theoretical model argues that retail sales per household are determined by three fundamental factors: the Market Environment (which is beyond the control of retail managers), Intertype Competition (which is influenced, but not controlled, by managers in each line of trade), and the Marketing Mix in each line of trade (which is set by managers). The essence of our argument is that the Market Environment determines a base level of sales per household; Intertype Competition may raise or lower sales in our focal lines of trade; and, the Marketing Mix in each line of trade can augment sales by (a) doing an above average job of appealing to customers and (b) countering the negative impact of Intertype Competition. Based on our research framework, we conduct a three-stage, hierarchical multiple regression analysis in each line of trade. Within Market Environment we include nine variables in a first-stage regression model: average number of people per household, household growth rate, average home size in square meters, income per household, population ratio aged 65+, auto ownership per household, distance to the prefectural capital city, residential land prices, and daytime population ratio. We expect each of these independent variables (except for population ratio aged 65+) to increase retail sales per household – which is our dependent variable. For Intertype Competition we use General Merchandise Stores (largely department stores and supercenters) that, in Japan, directly compete with Apparel, Food, and Furniture stores. In the second-stage regression model we include GMS sales per household along with the above nine Market Environment variables. GMS is anticipated to lower sales per household in Food stores, but is expected to raise sales per household in Apparel and Furniture stores as a spillover effect. For the Marketing Mix we measure four variables: assortment (proxied as square meters of selling space per store), service (employees per square meter of selling space), access (number of stores per land surface area of the city), and advertising (newspaper subscribers per household). These variables are included in the third-stage regression model along with the aforementioned ten independent variables; each of them should increase retail sales per household in its line of trade (e.g., the marketing mix for Food stores should only affect food sales per household). Thus, in of our analysis we show the results of eighteen regressions (i.e. the six census years and three lines of trade) . Our empirical research makes five contributions. First, we incorporate five independent variables that rarely (if ever) appear in studies of sales per household: out-shopping (daytime population ratio), home size, population ratio aged 65+, distance from the prefectural capital city, and residential land price. Second, we show the impact of intertype competition on sales in specific lines of retail trade. Third, we investigate data from five censuses that span a sixteen year period; few previous studies have examined changes in retail structure over such a lengthy time span. Fourth, we examine consumer goods retailers – who are an important target of urban planning and retail distribution policies of Japanese cities. Fifth, Japan had three important characteristics during the time span we examine: it was the world’s second largest economy and it is a nation of gradually declining population. As such, it may be a harbinger of the future of retailing in other large, wealthy economies. Additionally, Japan has rarely been the focus of retail trade studies.