As a movement within Judaism of the Second Temple period, the Christian sect originated with no distinct institutions, including church buildings. Following the death of Jesus, the disciples reportedly continued in Jewish temple piety and table fellowship “from house to house” (Acts 2:46, 5:42). There are also references in Acts to household conversions (Acts 16:15, 34; 18:8) in which the earliest preaching was conducted in the homes of individuals, who offered hospitality to Paul (or others). This basic picture is substantiated by the letters of Paul, the earliest writings of the New Testament (50–60 CE). Paul refers to entire households forming churches (1 Cor. 1:16, Rom 16:15), and he regularly addresses greetings in the letters to churches in the house of so-and-so (1 Cor. 16:19; Rom. 16:5; Phlm. 2; Col. 4:15). These references suggest that existing households served as a nucleus of organization with cellular communities formed around them. In some of Paul's communities—in large urban centers—it appears that there were several such house-church cells, often identified by the home of the patron. For example, in Corinth there were as many as five or six known groups during the 50s, and Paul seems to address eight distinct cells in Rome in his final salutations of the Roman letter (Rom. 16:3–16). The existence of several such cells organized around individual household social ties contributed to diversity in the worship, organization, and social life of early Christian groups.

The extended-household structure typical of the Hellenistic-Roman world included not only immediate family members, but also other relatives, domestic slaves, and a larger coterie of freedmen, hired workers, business associates, and other clients or dependents. The head of the household (Lat., pater familias) served as patron and host to the group that met in the home, reflecting similar patterns of patron-client social relations to that of the household structure. Paul regularly stayed in the homes of the house-church patron on his visits, who provided him with financial support (Rom. 16:23; Phlm. 18–22; Phil. 4:10–20). Increasingly in the Roman period, women were able to retain functional control of personal estates and manage their own households (mater familias). A number of Paul's house-church communities were under such patronesses (Nympha, Col. 4:15; Chloe, 1 Cor.1:11; Lydia, Acts16:14–16; Euodia and/or Syntyche, Phil. 4:2). The best example is Phoebe (Rom. 16:2), who is clearly called the patroness (Gk., prostatis) of a house church at Kenchreai (a port of Corinth) and was Paul's host and supporter. Her wealth and social standing are attested by the fact that she was the one who actually carried Paul's letter to the Christians of Rome; she was sent, at least in part, to pave the way for Paul's arrival among Rome's diverse house churches and to raise financial support for his planned mission to Spain. In other cases, the wife seems to be the predominant member of a couple who hosted the church in their house (cf. Prisca, Rom.16:3). The case of Prisca and Aquila indicates further that some of these individuals had substantial means: they were able to travel and set up house churches in Corinth, Ephesus, and Rome (Acts 18:2–4; 1 Cor. 16:19; Rom.16:3–5).

House Churches

HOUSE CHURCHES. Figure 1. Reconstruction of the Christian Building at Dura-Europos, before and after it was renovated into a house church. (After White, 1990, p. 108)

view larger image

The most common reference for the place of Christian meetings in the New Testament is simply to houses, but with little or no description of the actual type of edifice or particular area of activity. Other references in Paul's letters suggest that the communal worship met in the dining room of the house and around a common table, where the meal proper served as a focal point for the activities of worship. Other places used for meetings included a domestic “upper room” (Acts 20:7) and a commercial hall (Gk., scholē) of indeterminate plan (Acts 19:9). The apocryphal Acts of Paul (late second century) further suggests that Christians in Rome met in large warehouses (Lat., horrea; cf. Passio Pauli 1). That such private or semiprivate architecture was used by Christian groups is not in itself unique. Many other small or newly arrived religious groups and collegial associations in the fluid and mobile social environment of the Roman Empire regularly adapted houses or other private edifices as places of assembly or cultic activity. All six of the synagogue buildings of the Diaspora (second century BCE-fourth century CE) known from archaeological remains were adapted from existing edifices, and five of the six were originally houses or apartment buildings (Lat., insulae). Often, the adaptation of these structures progressed gradually through multiple stages of renovation and remodeling to accommodate specific religious or social functions. The cult of Mithras is similarly predominantly found to have adapted existing edifices for its purposes. Of more than sixty excavated Mithraic sanctuaries (Lat., mithraea) from across the Roman Empire, only ten were constructed de novo for the cult; the remainder were found in renovated houses, apartments, shops, warehouses, basement cryptoportici, and the like. This includes all the excavated mithraea in Ostia and Rome, as well as the one at Caesarea Maritima (Israel). [See Caesarea.] At Dura-Europos (Roman Syria/Mesopotamia), there are three such adapted houses on the same street: a synagogue, a mithraeum, and a Christian building (see figure 1). [See Dura-Europos; Synagogues.]

In archaeological terms, a distinction needs to be made between a house or other edifice that might be used for religious purposes on a casual basis and an edifice actually renovated architecturally to meet specific spatial or icono-graphic needs. While it is clear that in Paul's day (and through much of the second century) Christians met in the homes of members, it does not appear that significant renovations occurred. These cases are properly termed house churches, and they continued in primarily domestic or other nonreligious functions. Yet, it is not likely that any archaeological evidence can be discerned for them because they lack distinguishing architectural, spatial, or artistic features. Archaeological evidence arises only with physical adaptation of the edifice. While the terminology has been inconsistent, it is preferable to use the term domus ecclesiae (“house of the church”) to denote renovated Christian edifices on their way to becoming more formal church buildings. This terminology is reflected in some third-century Christian writers, where further expansion and remodeling are indicated. Such edifices are referred to by Eusebius (Hist. Eccl. 7.30.19, 8.1.5) in Greek as oikos ekklesias (= Lat., domus ecclesiae) or oikos kyriakou (“house of the Lord”).

The earliest and most clearly attested archaeological evidence of a house renovated architecturally into a place of Christian assembly and worship is from Dura-Europos. It was buried substantially intact in 256 CE, when the city was destroyed. It was a typical domestic edifice by local standards, with rooms grouped around a central court. While Christians might have been in Dura earlier, no other direct archaeological evidence exists of their activity or assembly. There is no direct evidence that the Dura building was used as a house church prior to its renovation (in c. 241 CE), though it is possible. When it was renovated all domestic functions ceased, and the building was given over completely to the spatial needs of religious (or cultic) activities. Even though the building retained its exterior domestic form and plan, it had become a “church building.” Such features can be detected archaeologically from the nature of the renovation and other surviving epigraphic and artistic decoration. On the south side of the court, a partition wall was removed in order to open up two rooms to form a rectangular hall for assembly. A small dais was then installed at the east end of the room. A small adjoining room seems to have been used for storage. A single fragmentary inscription preserved from the room likely refers to ecclesiastical personnel. In the northwest corner of the house another room had a small font pool with an arched canopy installed to serve as a baptistery. [See Baptisteries.] This room also shows clear Christian usage in its artistic program, which depicts scenes from the life of Jesus on the walls. Other memorial inscriptions from the walls of this room clearly reflect a Christian identity.

Several other sites from across the Roman world provide further archaeological evidence for the adaptation and renovation of existing buildings prior to the implementation of formal basilical church architecture under Constantine (c. 314–319). [See Basilicas; Churches.] The evidence suggests that partial adaptation occurred as an intermediate stage, and that subsequent enlargements or rebuilding overlaid earlier phases of the renovation of domestic or private architecture. One site, at the Lullingstone Villa, near Eynsford, England, has only part of the house given over for renovation as a kind of “chapel wing”; the site is important also for its relatively late date (c. 350–410), which shows that the process of gradual adaptation continued well after the empire had become officially Christian and public Christian architecture had begun. The continued renovation and reuse of sites over centuries has often made the archaeological evidence difficult to read. For example, in Rome, a large apartment building with shops on the ground floor and domestic quarters above seems to have been partially renovated for use as a Christian hall in the third century. This basic plan may have continued until the beginning of the fifth century before being fully rebuilt in basilical form; however, some of the earlier structure was preserved in the foundation levels below the basilica, now called the Church of Saints John and Paul (Ss. Giovanni e Paolo). Similarly, the Church of St. Clement (San Clemente) near the Colosseum was originally built in basilical form in about 400, over a large third century hall that may have been used by Christians. In turn, this hall stood over a large warehouse complex dating to the first century that Christian tradition associates with the name of Clement of Rome (d. about 96) and a church ostensibly located on his property. No direct archaeological evidence indicates such usage in the lower layers, even though the adjacent house clearly had a mithraeum installed in a lower level cortile. Similarly, there is no evidence that the so-called House of St. Peter at Capernaum was actually used as a house church, or domus ecclesiae, prior to the construction of a quadrilateral enclosure in the fourth century, later rebuilt on an octagonal plan in the fifth and sixth centuries. [See Capernaum.] In this case it is better to view the building in the context of the beginning of memorial architecture stimulated by Constantine to commemorate Christian holy sites, rather than its being in direct continuity with the older pattern of house church and domus ecclesiae.

Other archaeological evidence suggests that by the mid-third century Christians were beginning to enlarge spaces on a rectangular hall plan, but well before the beginning of formal basilical church architecture. Such buildings may be called aula ecclesiae, or “hall churches.” At least one case, from Parentium (Roman Istria, modern Parenzo or Porec), shows a private house first renovated as a domus ecclesiae (third century) and subsequently renovated to form a large hall (fourth century), before it was converted to the basilical plan (fifth-sixth centuries; now known as the Basilica Eufrasiana). A hall was also installed in a block of private houses in the early fourth century at Umm el-Jimal (in Roman Arabia, modern Jordan). [See Umm el-Jimal.] Other examples can be found in the first levels below the Church of San Chrysogono in Rome (c. 310), the double halls below the church at Aquileia (c. 317–319), a small plane hall from Qirk Bize in Syria (c. 320–333), and the first hall below the Octagonal Church of St. Paul at Philippi, Greece (c. 343). Each of these cases was transformed, by subsequent phases of renovation and construction, into more formal or recognizable church architecture (either basilical or octagonal in plan).


  • Finney, Paul Corby. “Early Christian Architecture: The Beginnings.” Harvard Theological Review 81 (1988): 319–339.
  • Judge, E. A. The Social Organization of Christian Groups in the First Century. London, 1960.
  • Klauck, Hans-Josef. Hausgemeinde und Hauskirche im frühen Christentum. Stuttgart, 1981.
  • Kraeling, Carl H. The Christian Building. The Excavations at Dura-Europos, Final Report 8, part 2. New Haven, 1967.
  • Krautheimer, Richard. Corpus basilicarum christianarum romae. 5 vols. Vatican City, 1939–1956.
  • Krautheimer, Richard Early Christian and Byzantine Architecture. 3d ed. Harmondsworth, 1981.
  • Malherbe, Abraham J. Social Aspects of Early Christianity. 2d ed. Philadelphia, 1983.
  • Meeks, Wayne A. The First Urban Christians: The Social World of the Apostle Paul. New Haven, 1983.
  • Petersen, Joan M. “House Churches in Rome.” Vigiliae Christianae 23 (1969): 264–272.
  • White, L. Michael. Building God's House in the Roman World: Architectural Adaptation among Pagans, Jews, and Christians. Baltimore, 1990.
  • White, L. Michael. The Social Origins of Christian Architecture, vol. 1, Architectural Adaptation among Pagans, Jews, and Christians (reprint of above); vol. 2 Texts and Monuments of the Christian Domus Ecclesiae in Its Environment. Minneapolis, 1995.

L. Michael White