Although population estimates are available for earlier years, the first official Portuguese census was taken in 1864. It showed a population of approximately 4.3 million (see table 2, Appendix). Thereafter, the population increased slowly at rates often well under 1 percent per year. Only during the 1930s and 1940s did the population increase at over 1 percent per year. During the 1960s, the population actually fell by over 300,000 and in 1970 amounted to more than 8.5 million. During the early 1970s, population continued to fall or was stagnant. This demographic trend was the result of widespread emigration. Many Portuguese left their country in these years to find employment abroad or to avoid military service in the wars Portugal was fighting in its colonies in Africa.

In 1974 the country's population showed its first sizeable increase and by 1981 reached nearly 9.8 million, 1.2 million more than it had been ten years earlier. The settling in Portugal of an estimated 500,000 to 800,000 refugees from the country's African colonies accounted for most of this increase. During the first half of the 1980s, the population grew at a rate of about 0.8 percent a year, then declined. As of the early 1990s, population growth was estimated at 0.4 percent a year. By the beginning of 1992, the population of Portugal, including the Azores and Madeira, was estimated at nearly 10.5 million. Population specialists projected that if existing trends continued, the country's population would peak at 10.8 million in 2010 and fall to 10.5 in 2025.

This population was not evenly distributed. As of the late 1980s, continental Portugal had an average population density of 109.6 persons per square kilometer, but some districts were much more crowded than others. The eastern districts bordering Spain, with the exception of Faro, had the lowest population density, ranging between 17.0 per square kilometer in Beja and 35.6 per square kilometer in Guarda. Coastal districts from the northern border down to and including Setúbal registered the highest concentrations of people. The districts of Lisbon and Porto, with 770.2 and 697.5 persons per square kilometer, respectively, were as densely populated as many urban regions of Northern Europe (see table 3, Appendix).

Some of these differences in population density resulted from topography. Mountainous regions typically contain fewer people than flat coastal regions. But some differences resulted from migration from one area to another within Portugal or from migration abroad. During the period 1911-89, five districts, all of them bordering Spain in the east, lost population: Guarda lost about one-fourth of its population, Beja and Castelo Branco lost about one-tenth, and Bragança and Portalegre each lost about onetwentieth . The only eastern district posting a gain in this period was Évora, which grew by about one-sixth. Two inland districts, Vila Real and Viseu, showed almost no growth another inland district, Santarém, with significantffd nt industrial employment, grew by one-half. All coastal districts gained in population during this period. Coimbra and Faro grew by onefourth , Aveiro and Braga doubled their populations, the districts of Lisbon and Porto increased by two-and-one-half times, and Setúbal increased more than three times. The Azores showed almost no gain in population, but that of Madeira grew by two-thirds.

The main areas of population growth were urban centers and the district capitals. The urban-industrial centers along the coast--Lisbon, Porto, Setúbal--took in the largest numbers of new immigrants. However, only the cities of Lisbon and Porto had significant populations, approximately 830,000 and 350,000, respectively, at the end of the 1980s. They were followed by Amadora with 96,000 (part of greater Lisbon), Setúbal with 78,000, and Coimbra with 75,000. At the beginning of the 1990s, therefore, some two-thirds of all Portuguese still lived in what were classified as rural areas despite the significant growth of some urban areas.

The Lisbon area was the region of greatest population growth in absolute terms, in part it was the seat of much of the country's governmental apparatus, as well as its manufacturing and service sector jobs. Until the 1960s, the area's population increases were mainly inside the city of Lisbon, but since then the suburbs have grown most rapidly. The central city's population remained largely stagnant or even declined in some years, while that of the suburbs surged. High city rents, crowding, the decline of old neighborhoods, pollution, and the squeezing out of housing by commercial enterprises were among the causes of this new suburbanization of Lisbon's outlying districts.

Government population estimates showed that in the late 1980s women outnumbered men by a wide margin and that the number of old persons in Portugal was unusually high. The 1864 census and every census since has shown that women outnumber men. In 1988 this was the case in all but two of the districts of continental Portugal, Beja and Bragança. The greatest disproportions were found in northern and central areas where male emigration was most intense. However, during the 1980s, men formed the majority in twenty-two of the country's 305 municipalities. Eighteen of these statistically unusual municipalities were in southern Portugal.

Portugal has long had an aging population. The percentage of the population under age thirty has been decreasing since 1900. Moreover, the rate at which the country's population has aged accelerated as ever more young Portuguese males in their physical prime left the country. Between 1960 and 1990, the percentage of those under fifteen fell from 29.0 to 20.9, while that of those sixty-five and older rose from 8.1 to 13.1. The north had a disproportionate number of old and very young people, mainly those still too young to migrate. In some areas of Portugal where employment has been available, this preponderance was not the case. Lisbon and the growth areas of Santarém and Setúbal had a disproportionate share of those of working age, between twenty and sixty-five (see fig. 6).

Data as of January 1993