Primate neoplastic and finite cell lines were tested in one in vivo and two in vitro test systems: adult nude mice, muscle organ culture (MOC) and soft agarose (SA). Comparison of the sensitivity of the systems indicated that nude mice were inferior to either in vitro system: WI-38 VA13 (an SV40 transformed cell line) did not cause tumours in these animals yet it behaved as if it were neoplastic in MOC and formed colonies in SA. There was complete correlation between results obtained in MOC and SA. All cell lines which produced tumors in vivo were positive in both in vitro test systems. None of the lines which showed normal patterns in MOC and in SA was tumorigenic in nude mice. Since testing in vitro is simpler, faster, and is thought to be reliable, we recommend SA followed by MOC as the initial assays for determining tumorigenicity of cells.