PIP: This paper uses map analysis to study the transition of family limitation practice in Taiwan between 1961-80. The innovation-diffusion perspective emphasizes that birth control, particularly contraception, is a recent innovation and is essentially new in human culture. The innovation-diffusion theory assumes that the decline of fertility began in a setting where there was no, or at most very limited, previous practice of birth control. The theory emphasizes the importance of the spread of information. It also assumes that innovation starts in metropolitan centers, diffuses to other urban places with some delay, and penetrates to rural areas still later. Innovation behavior also diffuses from 1 area to another which is culturally and linguistically similar. Although there was some urban to rural diffusion from the Taiwan family planning program, the government supported program provided services more evenly between urban and rural areas, thus somewhat limiting the diffusion effect from the program. For the diffusion of family practice in Taiwan, it is expected that the availability of of information about and means of family limitation practice may effect the rate of the increase of small m values -- an index of family limitation -- in an area. The case study of Pingtung county shows that the demand-side diffusion from urban to rural areas was important in the earlier decade of the transition of family plimitation practice, but distance from urban center was less important as practice became more uniform through diffusion. Ethnicity, whether or not the township was dominated by Hakka or Fukienese, also seems to have played an important role in determining the pace at which the local residents adopted family practice limitation. Hakka townships seem to have adopted family limitation practice more slowly than Fukienese townships about the same distance from the urban center. The map analysis of Pingtung county provides descriptive evidence to support the diffusion of family limitation from urban centers to distant areas, while ethnic variables like Hakka population tend to delay the adoption of family limitation practice. In general, the urban center had higher m values than the surrounding rural areas in Pingtung county and for areas other than the urban center the the level of m values is a negative function of the distance to the urban center.