A multilevel model of macro and micro variables is used to examine the fertility determinants in the 4 regions of Thailand. The level and pace of fertility decline has varied across regions. Fertility behavior is considered at parity 0, at parities 1-3, and among women with 4 children. Contextual level variables are the status of women, the value of children, infant mortality in 1970, and population pressure. Micro data were obtained from 1970 and 1980 census files for women 15-49 years matched with own children data (aged 1-4 years and born between 1975-79). In general, contextual variables were more important determinants at parity 1-3 and had no effect at parity 0, and individual differences were more important at parity 0. The findings in the regression analyses show consistent results with prior research: older ages effecting a decrease on fertility, migrants having fewer children, higher female education and higher occupational status of husbands lower fertility. When marriage status is controlled at parity 0 the relationship between individual variables and marriage shows sign changes and reduced effects. The effects of individual variables are strongest at parity 4 and over, but affect all parities. Variations between regions on individual variables are small. Contextual variables show less consistency in results. Status of women results show significant negative effects at parity 4 for the central and northern regions. At parity 1-3 the status of women with contextual variables shows status as an important determinant of fertility. Provinces with higher levels of education had lower fertility. The value of children is negatively associated with fertility. The effect is greatest for parity 1-3 women. The north shows significant relationships, while the other regions are inconsistent; the implications are discussed. Infant mortality is not related to fertility except at 0 parity in the north (positive relationship). Population pressure is related to lower fertility in the northeast and south; in the north the results are significant but the sign changes. Only with parity 4 in the south were family planning programs related to lower fertility (negative association); the results are positive in the north and northeast. The policy implications are that increasing the opportunities for women will have an impact on fertility, and fertility would be indirectly affected through improvements in women's status. Decentralization of policy making on fertility is needed along with awareness of the cultural and social environment.